Intelligent Design

What are the limits of Natural Selection? An interesting open discussion with Gordon Davisson

Spread the love

An interesting discussion, still absolutely open, has taken place in the last few days between Gordon Davisson and me on the thread:

What? Only an “extremely occasional” mutation is beneficial? But Darwinism… ?

Some very good friends, like Dionisio, Mung and Origenes, seem to have appreciated the discussion, which indeed has touched important issues. Origenes has also suggested that it could be transformed into an OP.

Well, I tought that it was probably a good idea, and luckily it did not require much work. 🙂   So, here it is. Gordon Davisson’s posts are in italics. It’s a bit long, and I am sorry for that!

I thank in advance Gordon Davisson for the extremely good contribution he has already given, and for any other contribution he will give. He is certainly invited to continue the discussion here, if he likes (and I do hope he does!). Of course, anyone else who could be interested is warmly invited to join.  🙂

Gordon Davisson (post #5):

Why is this supposed to be a problem for “Darwinism”? A low rate of beneficial mutations just means that adaptive evolution will be slow. Which it is.

And not as slow as it might appear, since the limiting rate is the rate of beneficial mutations over the entire population, not per individual. Although many beneficial mutations are wiped out by genetic drift before they have a chance to spread through the population, so that decreases the effective rate a bit. If I’ve accounted for everything the overall rate of fixation of beneficial mutations per generation should be: (fraction of mutations that’re beneficial) * (fraction of beneficial mutations that aren’t wiped out by genetic drift) * (# of mutations per individual) * (population).

Florabama’s description is exactly wrong. Beneficial mutations don’t have to happen “in a row”, they can happen entirely independently of each other, and spread independently via selection. You may be thinking of the argument from irreducible complexity, but that’s an argument that evolution depends on mutations that are only beneficial in combination, which is a different matter. (And FYI evolutionists dispute how much of a problem this actually is. But again, that’s another matter.)

gpuccio (post #11):

Gordon Davisson:

You say:

And not as slow as it might appear, since the limiting rate is the rate of beneficial mutations over the entire population, not per individual.

Yes, but any “beneficial” mutation that appears in one individual will have to expand to great part of the population, if NS has to have any role in lowering the probabilistic barriers.

That means that:

1) The “beneficial” mutation must not only be “beneficial” in a general sense, but it must already, as it is, confer a reproductive advantage to the individual clone where it was generated. And the reproductive advantage must be strong enough to significantly engage NS (against the non-mutated form, IOWs all the rest of the population), and so escape genetic drift. That is something! Can you really think of a pathway to some complex new protein, let’s say dynein, a pathway which can “find” hundreds of specific, highly conserved aminoacids in a proteins thousands of aminoacid long, whose function is absolutely linked to a very complex and global structure, a pathway where each single new mutation which changes one aminoacid at a time confers a reproductive advantage to the individual, by gradually increasing, one step at a time, the function of a protein which still does not exist?

If you can, I really admire your imagination.

2) Each of those “beneficial mutations” (non existing, IMO, but let’s suppose they can exist) has anyway to escape drift and be selected and expanded by NS, so that it is present in most, or all the population. That’s how the following mutation can have some vague probability to be added. That must happen for each single step.

While that is simply impossible, because those “stepwise” mutations simply do not exist and never will exist, even if we imagine that they exist the process requires certainly a lot of time.

Moreover, as the process seems not to leave any trace of itself in the proteomes we can observe today, because those functionally intermediate forms simply do not exist, we must believe that each time the expansion of the new trait, with its “precious” single aminoacid mutation, must be complete, because it seems that it can erase all tracks of the process itself.

So, simple imagination is not enough here: you really need blind faith in the impossible. Credo quia absurdum, or something like that.

Then you say:

Although many beneficial mutations are wiped out by genetic drift before they have a chance to spread through the population, so that decreases the effective rate a bit.

Absolutely! And it’s not a bit, it’s a lot.

If you look at the classic paper about rugged landscape:

http://journals.plos.org/ploso…..ne.0000096

you will see that the authors conclude that a starting library of 10^70 mutations would be necessary to find the wild-type form of the protein they studied by RM + NS. Just think about the implications of that simple fact.

You say:

Beneficial mutations don’t have to happen “in a row”, they can happen entirely independently of each other, and spread independently via selection.

Yes, but only if each individual mutation confers a strong enough reproductive advantage. That must be true for each single specific aminoacid position of each single new functional protein that appears in natural history. Do you really believe that? Do you really believe that each complex functional stricture can be deconstructed into simple steps, each conferring reproductive advantage? Do you believe that we can pass from “word” source code to “excel” source code by single byte variations (yes, I am generous here, because a single aminoacid has at most about 4 bits of information, not 8), each of them giving a better software which can be sold better than the previous version?

Maybe not even “credo quia absurdum” will suffice here. There are limits to the absurd that can be believed, after all!

You say:

You may be thinking of the argument from irreducible complexity, but that’s an argument that evolution depends on mutations that are only beneficial in combination, which is a different matter.

No, the argument of IC, as stated by Behe, is about functions which require the cooperation of many individual complex proteins. That is very common in biology.

The argument of functional complexity, instead, is about the necessity of having, in each single protein, all the functional information which is minimally necessary to give the function of the protein itself. How many AAs would that be, for example, for dynein? Or for the classic ATP synthase?

Here, the single functional element is so complex that it requires hundreds of specific aminoacids to be of any utility. If that single functional element also requires to work with other complex single elements to give the desired function (which is also the rule in biology), then the FC of the system is multiplied. That is the argument of IC, as stated by Behe. The argument for FC in a single functional structure is similar, but it is directly derived form the concept of CSI as stated by Dembski (and others before and after him).

And finally you say:

And FYI evolutionists dispute how much of a problem this actually is. But again, that’s another matter.

It’s not another matter. It’s simply a wrong matter.

Both FC and IC are huge problems for any attempt to defend the neo-darwinian theory. I am not surprised at all that “evolutionists” dispute that, however. See Tertullian’s quote above!

Gordon Davisson (post #35):

Hi, gpuccio. Sorry about my late reply (as usual, I’m afraid). Before I comment specifically to what you said, I need to make a general comment that I still don’t see how the original point — that beneficial mutations are rare — refutes evolution. The arguments you’re making against evolution’s ability to create complex functional systems don’t seem to have a very close connection to the rate of beneficial mutations. Note that all of these would be considered beneficial mutations:

* Minor changes to an existing functional thing (protein, regulatory region, etc) that improve its function slightly.
* Minor changes to an existing functional thing that change its function slightly, in a way that makes it fit the organism’s current environment better.
* Changes that decrease function of something that’s overdoing its role (e.g. the mutation discussed here, which winds up giving people unusually strong bones).
* Mutations that create new functional systems.
* Mutations that are partway along a path to new functional systems, and are beneficial by themselves.

Your argument is (if I may oversimplify it a bit) essentially that the last two are vanishingly rare. But when we look at the overall rate of beneficial mutations, they’re mixed in with other sorts of beneficial mutations that’re completely irrelevant to what you’re talking about! Additionally, several types of mutations that’re critical in your argument are not immediately beneficial aren’t going to be counted in the beneficial mutation rate:

* Mutations that move closer to a new functional system (or higher-functioning version of an existing system), but aren’t actually there yet.
* Mutations that produce new functional systems that don’t immediately contribute to fitness.

Furthermore, one of the reasons for the rate of beneficial mutations may be low is that there may simply not be much room for improvement. For example, the experiment you cited about evolution on a rugged fitness landscape suggests that the wild-type version of the protein they studied may be optimal — it cannot be improved, whether by evolution or intelligent design or whatever. If that’s correct, the rate of beneficial mutations to this protein will be exactly zero, but that’s not because of any limitation of what mutations can do.

Now, on to your actual argument:

And not as slow as it might appear, since the limiting rate is the rate of beneficial mutations over the entire population, not per individual.

Yes, but any “beneficial” mutation that appears in one individual will have to expand to great part of the population, if NS has to have any role in lowering the probabilistic barriers.

That means that:

1) The “beneficial” mutation must not only be “beneficial” in a general sense, but it must already, as it is, confer a reproductive advantage to the individual clone where it was generated. And the reproductive advantage must be strong enough to significantly engage NS (against the non-mutated form, IOWs all the rest of the population), and so escape genetic drift. That is something!

I’d disagree slightly here. There isn’t any particular “strong enough” threshold; the probability that a beneficial mutation will “escape genetic drift” is roughly proportional to how beneficial it is. Mutations that’re only slightly beneficial thus become fixed at a lower (but still nonzero) rate.

Can you really think of a pathway to some complex new protein, let’s say dynein, a pathway which can “find” hundreds of specific, highly conserved aminoacids in a proteins thousands of aminoacid long, whose function is absolutely linked to a very complex and global structure, a pathway where each single new mutation which changes one aminoacid at a time confers a reproductive advantage to the individual, by gradually increasing, one step at a time, the function of a protein which still does not exist?

If you can, I really admire your imagination.

I’ll discuss some of these points more below, but just two quick things here: first, this is just an argument from incredulity, not an argument from actual knowledge or evidence. Second, the article you cited about a rugged fitness landscape showed that they were able to evolve a new functional protein starting from a random polypeptide (the limit they ran into wasn’t getting it to function, but in optimizing that function).

2) Each of those “beneficial mutations” (non existing, IMO, but let’s suppose they can exist) has anyway to escape drift and be selected and expanded by NS, so that it is present in most, or all the population. That’s how the following mutation can have some vague probability to be added. That must happen for each single step.

While that is simply impossible, because those “stepwise” mutations simply do not exist and never will exist, even if we imagine that they exist the process requires certainly a lot of time.

This is simply wrong. Take the evolution of atovaquone resistance in P. falciparum (the malaria parasite). Unless I’m completely misreading the diagram Larry Moran gives in http://sandwalk.blogspot.com/2…..ution.html, one of the resistant variants (labelled “K1”) required 7 mutations in a fairly specific sequence, and at most 4 of them were beneficial. In order for this variant to evolve (which it did), it had to pass at least 3 steps unassisted by selection (which you claim here is impossible) and all 4 beneficial mutations had to overcome genetic drift.

At least in this case, beneficial intermediates are neither as rare nor as necessary as you claim.

Moreover, as the process seems not to leave any trace of itself in the proteomes we can observe today, because those functionally intermediate forms simply do not exist, we must believe that each time the expansion of the new trait, with its “precious” single aminoacid mutation, must be complete, because it seems that it can erase all tracks of the process itself.

So, simple imagination is not enough here: you really need blind faith in the impossible. Credo quia absurdum, or something like that.

Except we sometimes do find such traces. In the case of atovaquone resistance, many of the intermediates were found in the wild. For another example, in https://uncommondescent.com/intelligent-design/double-debunking-glenn-williamson-on-human-chimp-dna-similarity-and-genes-unique-to-human-beings/, VJTorley found that supposedly-novel genes in the human genome actually have very near matches in the chimp genome.

Then you say:

Although many beneficial mutations are wiped out by genetic drift before they have a chance to spread through the population, so that decreases the effective rate a bit.

Absolutely! And it’s not a bit, it’s a lot.

If you look at the classic paper about rugged landscape:

http://journals.plos.org/ploso…..ne.0000096

you will see that the authors conclude that a starting library of 10^70 mutations would be necessary to find the wild-type form of the protein they studied by RM + NS. Just think about the implications of that simple fact.

That’s not exactly what they say. Here’s the relevant paragraph of the paper (with my emphasis added):

The question remains regarding how large a population is required to reach the fitness of the wild-type phage. The relative fitness of the wild-type phage, or rather the native D2 domain, is almost equivalent to the global peak of the fitness landscape. By extrapolation, we estimated that adaptive walking requires a library size of 10^70 with 35 substitutions to reach comparable fitness. Such a huge search is impractical and implies that evolution of the wild-type phage must have involved not only random substitutions but also other mechanisms, such as homologous recombination. Recombination among neutral or surviving entities may suppress negative mutations and thus escape from mutation-selection-drift balance. Although the importance of recombination or DNA shuffling has been suggested [30], we did not include such mechanisms for the sake of simplicity. However, the obtained landscape structure is unaffected by the involvement of recombination mutation although it may affect the speed of search in the sequence space.

In other words, they used a simplified model of evolution that didn’t include all actual mechanisms, and they think it likely that’s why their model says the wild type couldn’t have evolved with a reasonable population size. So it must’ve been intelligent design… or maybe just homologous recombination. Or some other evolutionary mechanism they didn’t include.

Or their model of the fitness landscape might not be completely accurate. I’m far from an expert on the subject, but from my read of the paper:

* They measured how much infectivity (function) they got vs. population size (larger populations evolved higher infectivity before stagnating), fit their results to a theoretical model of the fitness landscape, and used that to extrapolate to the peak possible infectivity … which matched closely to that of the wild type. But their experimental results only measured relative infectivities between 0.0 and 0.52 (using a normalized logarithmic scale), and the extrapolation from 0.52 to 1.0 is purely theoretical. How well does reality match the theoretical model in the region they didn’t measure?

* But it’s worse than that, because their measurements were made on one functional “mountain”, and the wild type appears to reside on a different mountain. Do both mountains have the same ruggedness and peak infectivity? They’re not only extrapolating from the base of a mountain to its peak, but from the base of one mountain to the peak of another. The fact that the infectivity of the wild type matches closely with their theoretical extrapolation of the peak is suggestive, but hardly solid evidence.

So between the limitations of their simulation of actual evolutionary processes and the limitations of the region of the landscape over which they gathered data, I don’t see how you can draw any particularly solid conclusions from that study.

Well, except that there are some conclusions available from the region of the landscape that they did make measurements on: between random sequences and partial function. They say:

The landscape structure has a number of implications for initial functional evolution of proteins and for molecular evolutionary engineering. First, the smooth surface of the mountainous structure from the foot to at least a relative fitness of 0.4 means that it is possible for most random or primordial sequences to evolve with relative ease up to the middle region of the fitness landscape by adaptive walking with only single substitutions. In fact, in addition to infectivity, we have succeeded in evolving esterase activity from ten arbitrarily chosen initial random sequences [17]. Thus, the primordial functional evolution of proteins may have proceeded from a population with only a small degree of sequence diversity.

This seems to directly refute your claim that stepwise-beneficial mutations cannot produce functional proteins. They showed that it can. And they also showed that (as with the atovaquone resistance example) evolution doesn’t require stepwise-beneficial paths either. They found that stepwise-beneficial paths existed up to a relative fitness of 0.4, but they experimentally achieved relative fitnesses up to 0.52! So even with the small populations and limited evolutionary mechanisms they used, they showed it was possible to evolve significantly past the limits of stepwise-beneficial paths.

I don’t have to imagine this. They saw it happen.

gpuccio (posts 36 -39, 41, 46, 48):

 

Gordon Davisson:

First of all, thank you for your detailed and interesting comments to what I wrote. You raise many important issues that deserve in depth discussion.

I will try to make my points in order, and I will split them in a few different posts:

1) The relevance of the rate of “beneficial” mutations.

You say:

Before I comment specifically to what you said, I need to make a general comment that I still don’t see how the original point — that beneficial mutations are rare — refutes evolution. The arguments you’re making against evolution’s ability to create complex functional systems don’t seem to have a very close connection to the rate of beneficial mutations.

I don’t agree. As you certainly know, the whole point of ID is to evaluate the probabilistic barriers that make it impossible for the proposed mechanism of RV + NS to generate new complex functional information. The proposed mechanism relies critically on NS to overcome those barriers, therefore it is critical to understand quantitatively how often RV occurs that can be naturally selected, expanded and fixed.

Without NS, it is absolutely obvious that RV cannot generate anything of importance. Therefore, it is essential to understand and demonstrate how much NS can have a role in modifying that obvious fact, and the rate of naturally selectable mutations (not of “beneficial mutations, because a beneficial mutation which cannot be selected because it does not confer a sufficient reproductive advantage is of no use for the model) is of fundamental importance in the discussion.

2) Types of “beneficial” mutations (part 1).

You list 5 types of beneficial mutations. Let’s consider the first 3 types:

Note that all of these would be considered beneficial mutations:
* Minor changes to an existing functional thing (protein, regulatory region, etc) that improve its function slightly.
* Minor changes to an existing functional thing that change its function slightly, in a way that makes it fit the organism’s current environment better.
* Changes that decrease function of something that’s overdoing its role (e.g. the mutation discussed here, which winds up giving people unusually strong bones).

Well, I would say that these three groups have two things in common:

a) They are mutations which change the functional efficiency (or inefficiency) of a specific function that already exists (IOWs, no new function is generated).

b) The change is a minor change (IOWs, it does not imply any new complex functional information).

OK, I am happy to agree that, however common “beneficial” mutations may be, they almost always, if not always, are of this type. that’s what we call “microevolution”. It exists, and nobody has ever denied that. Simple antibiotic resistance has always been a very good example of that.

Of course, while ID does not deny microevolution, ID theory definitely shows its limits. They are:

a) As no new function is generated, this kind of variation can only tweak existing functions.

b) While the changes are minor, they can accumulate, especially under very strong selective pressure, like in the case of antibiotic resistance (including malaria resistance). But gradual accumulation of this kind of tweaking takes long times even under extremely strong pressure, requires a continuous tweaking pathway that is not always existing, and is limited, however, by how much the existing function can be optimized by simple stepwise mutations.

I will say more about those points when I answer about malaria resistance and the rugged landscape experiment. I would already state here, however, that both those scenarios, that you quote in your discussion, are of this kind, IOWs they fall under one of these three definitions of “beneficial” mutations.

3) Types of “beneficial” mutations (part 2).

The last two types are, according to what you say:

* Mutations that create new functional systems.
* Mutations that are partway along a path to new functional systems, and are beneficial by themselves.

These are exactly those kinds of “beneficial” mutations that do not exist.

Let’s say for the moment that we have no example at all of them.

For the first type,are you suggesting that there are simple mutations that “create new functional systems”? Well, let’s add an important word:

“create new complex functional systems”?

That word is important, because, as you certainly know, the whole point of ID is not about function, but about complex function. Nobody has ever denied that simple function can arise by random variation.

So, for this type, I insist: what examples do you have?

You may say that even if you have no examples, it’s my burden to show that it is impossible.

But that is wrong. You have to show not only that it is possible, but that it really happens and has real relevance to the problem we are discussing. We are making empirical science here, not philosophy. Only ideas supported by facts count. So, please, give the facts.

I would say that there is absolutely no reason to believe that a “simple” variation can generate “new complex functional systems”. There is no example of that in any complex system. Can the change of a letter generate a new novel? Can the change of a byte generate a new complex software, with new complex functions? Can a mutation of 1 – 2 aminoacids generate a new complex biological system?

The answer is no, but if you believe differently, you are welcome: just give facts.

In the last type of beneficial mutations, you hypothesize, if I understand you well, that a mutation can be part of the pathway to a new complex functional system, which still does not exist, but can be selected because it is otherwise beneficial.

So, let’s apply that to the generation of a new functional protein, like ATP synthase. Let’s say the beta chain of it, which, as we all know, has hundreds of specific aminoacid positions, conserved from bacteria to humans (334 identities between E. coli and humans).

Now, what you are saying is that we can in principle deconstruct those 334 AA values into a sequence of 334 single mutations, or if you prefer 167 two AAs mutations, each of which is selected not because the new protein is there and works, but because the intermediate state has some other selectable function?

Well, I say that such an assumption is not reasonable at all. I see no logical reason why that should be possible. If you think differently, please give facts.

I will say it again; the simple idea that new complex functions can be deconstructed into simple steps, each of them selectable for some not specified reason, is pure imagination. If you have facts, please give them, otherwise that idea has not relevance in a scientific discussion.

4) Other types of mutation?

You add two further variations in your list of mutations. Here they are:

* Mutations that move closer to a new functional system (or higher-functioning version of an existing system), but aren’t actually there yet.
* Mutations that produce new functional systems that don’t immediately contribute to fitness.

I am not sure that I understand what you mean. If I understand correctly, you are saying that there are mutations which in the end will be useful, bur for the moment are not useful.

But, then, they cannot be selected as such. Do you realize what that means?

It means that they can certainly occur, but they have exactly the same probability to occur as any other mutation. Moreover, as they are no selected, they remain confined to the original individual or clone, unless they are fixed by genetic drift.

But again, they have exactly the same probability as any other mutation to be fixed by genetic drift.

That brings us to a very strong conclusion that is often overlooked by darwinists, especially the neutralists:

Any mutation that does not have the power to be naturally selected is completely irrelevant in regard to the probabilistic barriers because its probability is exactly the same as any other mutation to occur or to be fixed by drift.

IOWs, only mutations that can be naturally selected change the game in regard to the computation of the probabilistic barriers. Nothing else. All variation which cannot be naturally selected is irrelevant, because it is just a new random state, and is already considered when we compute the probabilities for a random search to get the target.

5) Optimal proteins?

You say:

Furthermore, one of the reasons for the rate of beneficial mutations may be low is that there may simply not be much room for improvement. For example, the experiment you cited about evolution on a rugged fitness landscape suggests that the wild-type version of the protein they studied may be optimal — it cannot be improved, whether by evolution or intelligent design or whatever. If that’s correct, the rate of beneficial mutations to this protein will be exactly zero, but that’s not because of any limitation of what mutations can do.

OK, I can partially agree. The proteins as we see them now are certainly optimal in most cases. But they were apparently optimal just from the beginning.

For example, our beloved ATP synthase beta chain already had most of its functional information in LUCA, according to what we can infer from homologies. And, as I have shown in my OPs about the evolution of information in vertebrates, millions of bits of new functional information have appeared at the start of the vertebrate branch, rather suddenly, and then remained the same for 400+ million years of natural history. So, I am not sure that the optimal state of protein sequences is any help for neo-darwinism.

Moreover, I should remind you that protein coding genes are only a very small part of genomes. Non coding DNA, which according to darwinists is mostly useless, can certainly provide ample space for beneficial mutations to occur.

But I will come back to that point in the further discussion.

I would like to specify that my argument here is not to determine how common exactly are beneficial mutations in absolute, but rather to show that rare beneficial mutations are certainly a problem for neo-darwinism, a very big problem indeed, especially considering that (almost) all the examples we know of are examples of micro-evolution, and do not generate any new complex functional information.

5) The threshold for selectability.

You say:

I’d disagree slightly here. There isn’t any particular “strong enough” threshold; the probability that a beneficial mutation will “escape genetic drift” is roughly proportional to how beneficial it is. Mutations that’re only slightly beneficial thus become fixed at a lower (but still nonzero) rate.

I don’t think we disagree here. Let’s say that very low reproductive advantages will not be empirically relevant, because they will not significantly raise the probability of fixation above the generic one from genetic drift.

On the other hand, even if there is a higher probability of fixation, the lower it is, the lower will be the effect on probabilistic barriers. Therefore, only a significant reproductive advantage will really lower the probabilistic barriers in a relevant way.

6) The argument from incredulity.

You say:

I’ll discuss some of these points more below, but just two quick things here: first, this is just an argument from incredulity, not an argument from actual knowledge or evidence. Second, the article you cited about a rugged fitness landscape showed that they were able to evolve a new functional protein starting from a random polypeptide (the limit they ran into wasn’t getting it to function, but in optimizing that function).

I really don’t understand this misuse of the “argument from incredulity” issue (are, of course, not the only one to use it improperly).

The scenario is very simple: in science, I definitely am incredulous of any explanation which is not reasonable, has no explanatory power, and especially is not supported by any fact.

This is what science is. I am not a skeptic (I definitely hate that word), but I am not a credulous person who believes in things only because others believe in them.

You can state any possible theory in science. Some of them will be logically inconsistent, and we can reject from the start. But others will be logically possible, but unsupported by observed facts and by sound reasoning. We have the right and the duty to ignore those theories as devoid of any true scientific interest.

This is healthy incredulity. The opposite of blind faith.

I will discuss the rugged landscape issue in detail, later.

7) Malaria resistance.

In the end, the only facts you provide in favour of the neo-darwinist scenario are those about malaria resistance and the rugged landscape experiment. I will deal with the first here, and with the second in next post.

You say:

This is simply wrong. Take the evolution of atovaquone resistance in P. falciparum (the malaria parasite). Unless I’m completely misreading the diagram Larry Moran gives in http://sandwalk.blogspot.com/2…..ution.html, one of the resistant variants (labelled “K1”) required 7 mutations in a fairly specific sequence, and at most 4 of them were beneficial. In order for this variant to evolve (which it did), it had to pass at least 3 steps unassisted by selection (which you claim here is impossible) and all 4 beneficial mutations had to overcome genetic drift.

At least in this case, beneficial intermediates are neither as rare nor as necessary as you claim.

Now, let’s clarify. In brief, my point is that malaria resistance, like simple antibiotic resistance in general, is one of the few known cases of microevolution.

As I have already argued in my post #36, microevolutionary events are characterized by the following:

a) No new function is generated, but only a tweaking of some existing function.

b) The changes are minor. Even if more than one mutation accumulates, the total functional information added is always small.

I will discuss those two points for malaria resistance in the next point, but I want to clarify immediately that you are equivocating what I wrote when you say:

This is simply wrong.

Indeed, you quote my point 2) from post #11:

“2) Each of those “beneficial mutations” (non existing, IMO, but let’s suppose they can exist) has anyway to escape drift and be selected and expanded by NS, so that it is present in most, or all the population. That’s how the following mutation can have some vague probability to be added. That must happen for each single step.”

But you don’t quote the premise, in point 1:

“1) The “beneficial” mutation must not only be “beneficial” in a general sense, but it must already, as it is, confer a reproductive advantage to the individual clone where it was generated. And the reproductive advantage must be strong enough to significantly engage NS (against the non-mutated form, IOWs all the rest of the population), and so escape genetic drift. That is something! Can you really think of a pathway to some complex new protein, let’s say dynein, a pathway which can “find” hundreds of specific, highly conserved aminoacids in a proteins thousands of aminoacid long, whose function is absolutely linked to a very complex and global structure, a pathway where each single new mutation which changes one aminoacid at a time confers a reproductive advantage to the individual, by gradually increasing, one step at a time, the function of a protein which still does not exist?

I have emphasized the relevant part, that you seem to have ignored. Point 2 is referring to that scenario.

It is rather clear that I am speaking of the generation of bew complex functional information, and I even make an example, dynein.

So, I am not saying that no beneficial mutation can be selected, or that when that happens, like in microevolution, we cannot find the intermediate states.

What I am saying is that such a model cannot be applied to the generation of new complex final information, like dynein, because it is impossible to decosntruct a new complex functional unit into simple steps, each of them naturally selectable, while the new protein still does not even exist.

So, what I say is not wrong at all, and mt challenge to imagine such a pathway for dynein, of for ATP synthase beta chain, or for any of the complex functional proteins that appear in the course of natural history, or to find intermediates of that pathway, remains valid.

But let’s go to malaria.

I have read the Moran page, and I am not sure of your interpretation that 7 mutations (4 + 3) are necessary to give the resistance. Indeed, Moran says:

“It takes at least four sequential steps with one mutation becoming established in the population before another one occurs.”

But the point here is not if 4 or 7 mutations are needed. The point is that this is a clear example of microevolution, although probably one of the most complex that have been observed.

Indeed:

a) There is no generation of a new complex function. Indeed, there is no generation of a new function at all, unless you consider becoming resistant to an antibiotic because a gene loses the function to uptake the antibiotic a new “function”. Of course, we can define function as we like, but the simple fact is that here there is an useful loss of function, what Behe calls “burning the bridges to prevent the enemy from coming in”.

b) Whatever out definition of function, the change here is small. It is small if it amounts to 4 AAs (16 bits at most), it is small if it amounts to 7 aminoacids (28 bits at most).

OK, I understand that Behe puts the edge to two AAs in his book. Axe speaks of 4, from another point of view.

Whatever. The edge is certainly thereabout.

When I have proposed a threshold of functional complexity to infer design for biological objects, I have proposed 120 bits. That’s about 35 AAs.

Again, we must remember that all known microevolutionary events have in common a very favourable context which makes optimization easier:

a) They happen in rapidly reproducing populations.

b) They happen under extreme environmental pressure (the antibiotic)

c) The function is already present and it can be gradually optimized (or, like in the case of resistance, lost).

d) Only a few bits of informational change are enough to optimize or lose the function.

None of that applies to the generation of new complex functional information, where the function does not exist, the changes are informationally huge, and environmental pressure is reasonably much less than reproducing under the effect of a powerful antibiotic.

8) VJ’s point:

You say:

VJTorley found that supposedly-novel genes in the human genome actually have very near matches in the chimp genome.

It’s funny that you quote a point that I consider a very strong argument for ID.

First of all, VJ’s arguments are in confutation of some statements by Cornelius Hunter, with whom I often disagree.

Second, I am not sure that ZNF843 is a good example, because I blasted the human protein and found some protein homologs in primates, with high homology.

Third, there are however a few known human proteins which have no protein counterpart in other primates, as VJ correctly states. These seem to have very good counterparts in non coding DNA of primates.

So, if we accept these proteins as real and functional (unfortunately not much is known about them, as far as I know), then what seems to happen is that:

a) The sequence appears in some way in primates as a non coding sequence. That means that no NS for the sequence as representing a protein can take place.

b) In some way, the sequence acquires a transcription start in humans, and becomes an ORF. So the protein appears for the first time in humans and, if we accept the initial assumption, it is functional.

Well, if that kind of process will be confirmed, it will be a very strong evidence of design. the sequence is prepared in primates, where is seems to have no function at all, and is activated in humans, when needed.

The origin of functional proteins from non coding DNA, which is gaining recognition in the recent years, is definitive evidence of design. NS cannot operate on non coding sequences, least of all make them good protein coding genes. So, the darwinian mechanism is out, in this case.

9) The rugged landscape experiment

OK, this is probably the most interesting part.

For the convenience of anyone who may be reading this, I give the link to the paper:

http://journals.plos.org/ploso…..=printable

First of all, I think we can assume, for the following discussion, that the wild-type version of the protein they studied is probably optimal, as you suggested yourself. In any case, it is certainly the most functional version of the protein that we know of.

Now, let’s try to understand what this protein is, and how the experiment was realized.

The protein is:

G3P_BPFD (P03661).

Length: 424 AAs.

Funtion (from Uniprot):

“Plays essential roles both in the penetration of the viral genome into the bacterial host via pilus retraction and in the extrusion process. During the initial step of infection, G3P mediates adsorption of the phage to its primary receptor, the tip of host F-pilus. Subsequent interaction with the host entry receptor tolA induces penetration of the viral DNA into the host cytoplasm. In the extrusion process, G3P mediates the release of the membrane-anchored virion from the cell via its C-terminal domain”

I quote from the paper:

Infection of Escherichia coli by the coliphage fd is mediated by the minor coat protein g3p [21,22], which consists of three distinct domains connected via flexible glycine-rich linker sequences [22]. One of the three domains, D2, located between the N-terminal D1 and C-terminal D3 domains, functions in the absorption of g3p to the tip of the host F-pilus at the initial stage of the infection process [21,22]. We produced a defective phage, ‘‘fdRP,’’ by replacing the D2 domain of the fd-tet phage with a soluble random polypeptide, ‘‘RP3-42,’’ consisting of 139 amino acids [23].

So, just to be clear:

1) The whole protein is implied in infectivity

2) Only the central domain has been replaced by random sequences

So, what happens?

From the paper:

The initial defective phage fd-RP showed little infectivity, indicating that the random polypeptide RP3-42 contributes little to infectivity.

Now, infectivity (fitness) was measured by an exponential scale, in particular as:

W = ln(CFU) (CFU = colony forming units/ml)

As we can see in Fig. 2, the fitness of the mutated phage (fd-RP) is 5, that is:

CFU = about 148 (e^5)

Now, always from Fig 2 we can see that the fitness of the wildtype protein is about 22.5, that is:

CFU = about 4.8 billions

So, the random replacement of the D2 domain certainly reduces infectivity a lot, and it is perfectly correct to say that the fd-RP phage “showed little infectivity”.

Indeed, infectivity has been reduced of about 32.6 million times!

But still, it is there: the phage is still infective.

What has happened is that by replacing part of the g3p protein with random sequences, we have “damaged” the protein, but not to the point of erasing completely its function. The protein is still there, and in some way it can still work, even with the have damage/deformation induced by our replacement.

IOWs, the experiment is about retrieving an existing function which has been artificially reduced, but not erased. No new function is generated, but an existing reduced function is tweaked to retrieve as much as possible of its original functionality.

This is an important point, because the experiment is indeed one of the best contexts to measure the power of RM + NS in the most favorable conditions:

a) The function is already there.

b) Only part of the protein has been altered

c) Phages are obviously a very good substrate for NS

d) The environmental pressure is huge and directly linked to reproductive success (a phage which loses infectivity cannot simply reproduce).

IOWs, we are in a context where NS shoul really operate at its most.

Now, what happens?

OK, some infectivity is retrieved by RM. How much?

At the maximum of success, and using the most numerous library of mutations, the retrieved infectivity is about 14.7 (see again Fig. 2). Then the adaptive walk stops.

Now, that is a good result, and the authors are certainly proud of it, but please don’t be fooled by the logarithmic scale.

An infectivity of 14.7 corresponds to:

about 2.4 million CFU

So, we have an increase of:

about 17000 times as stated by the authors.

But, as stated by the authors, the fitmess should still increase of about 2000 times (fitness 7.6) to reach the functionality of the wild type. that means passing from:

2.4 million CFU

to

4.8 billion CFU

So, even if some good infectivity has been retrieved, we are still 2000 times lower than the value in the wild type!

And that’s the best they could achieve.

Now, why that limit?

The authors explain that the main reason for that is the rugged landscape of protein function. That means that RM and NS achieve some good tweaking of the function, but starting from different local optima in the landscape, and those local optima can go only that far.

The local optimum corresponding to the wildtype has never been found. See the paper:

The sequence selected finally at the 20th generation has ~W = 0.52 but showed no homology to the wild-type D2 domain, which was located around the fitness of the global peak. The two sequences would show significant homology around 52% if they were located on the same mountain. Therefore, they seem to have climbed up different mountains

The authors conclude that:

The question remains regarding how large a population is required to reach the fitness of the wild-type phage. The relative fitness of the wild-type phage, or rather the native D2 domain, is almost equivalent to the global peak of the fitness landscape. By extrapolation, we estimated that adaptive walking requires a library size of 10^70 with 35 substitutions to reach comparable fitness. Such a huge search is impractical and implies that evolution of the wildtype phage must have involved not only random substitutions but also other mechanisms, such as homologous recombination.

Now, having tried to describe in some detail the experiment itself, I will address your comments.

10) Your comments about the rugged landscape paper

You say:

That’s not exactly what they say. Here’s the relevant paragraph of the paper (with my emphasis added):

But it is exactly what they say!

Let’s see what I wrote:

“you will see that the authors conclude that a starting library of 10^70 mutations would be necessary to find the wild-type form of the protein they studied by RM + NS.

(emphasis added)

Now let’s see what they said:

By extrapolation, we estimated that adaptive walking requires a library size of 10^70 with 35 substitutions to reach comparable fitness. Such a huge search is impractical and implies that evolution of the wild-type phage must have involved not only random substitutions but also other mechanisms, such as homologous recombination.

(I have kept your emphasis).

So, the point is that, according to the authors, a library of 10^70 sequences would be necessary to find the wildtype by random substitutions only (plus, I suppose, NS).

That’s exactly what I said. Therefore, your comment, that “That’s not exactly what they say” is simply wrong.

Let’s clarify better: 10^70 is a probabilistic resource that is beyond the reach not only of our brilliant researchers, but of nature itself!

It seems that your point is that they also add that, given that “such a huge search is impractical” (what a politically correct adjective here! ), that should:

“imply that evolution of the wild-type phage must have involved not only random substitutions but also other mechanisms, such as homologous recombination.”

which is the part you emphasized.

As if I had purposefully left out such a clarifying statement!

Well, of course I have purposefully left out such a clarifying statement, but not because I was quote-mining, but simply because it is really pitiful and irrelevant. Let’s say that I wanted to be courteous to the authors, who have written a very good paper, with honest conclusions, and only in the end had to pay some minimal tribute to the official ideology.

You see, when you write a paper, and draw the conclusions, you are taking responsibilities: you have to be honest, and to state only what can be reasonably derived from the facts you have given.

And indeed the authors do that! They correctly draw the strong conclusion that, according to their data, RM + NS only cannot find the wildtype in their experiment (IOWs, the real, optimal function), unless we can provide a starting library of 10^70 sequences, which, as said, is beyond the reach of nature itself, at least on our planet. IOWs, let’s say that it would be “impractical”.

OK, that’s the correct conclusion according to their data. They should have stopped here.

But no, they cannot simply do that! So they add that such a result:

implies that evolution of the wild-type phage must have involved not only random substitutions but also other mechanisms, such as homologous recombination.

Well, what is that statement? Just an act of blind faith in neo-darwinism, which must be true even when facts falsify it.

Is it a conclusion derived in any way from the data they presented?

Absolutely not! There is nothing in their data that suggests such a conclusion. They did not test recombination, or other mechanisms, and therefore they can say absolutely nothing about what it can or cannot do. Moreover, they don’t even offer any real support from the literature for that statement. They just quote one single paper, saying that “the importance of recombination or DNA shuffling has been suggested”. And yet they go well beyond a suggestion, they say that their “conclusion” is implied. IOWs logically necessary.

What a pity! What a betrayal of scientific attitude.

If they really needed to pay homage to the dogma, they could have just said something like “it could be possible, perhaps, that recombination helps”. But “imply”? Wow!

But I must say that you too take some serious responsibility in debating that point. Indeed, you say:

In other words, they used a simplified model of evolution that didn’t include all actual mechanisms, and they think it likely that’s why their model says the wild type couldn’t have evolved with a reasonable population size. So it must’ve been intelligent design… or maybe just homologous recombination. Or some other evolutionary mechanism they didn’t include.

Well, they didn’t use “a simplified model of evolution”. They tested the official model: RM + NS. And it failed!

Since it failed, they must offer some escape. Of course, some imaginary escape, completely unsupported by any facts.

But the failure of RM + NS, that is supported by facts, definitely!

I would add that I cannot see how one can think that recombination can work any miracle here: after all, the authors themselves have said that the local optimum of the wildtype has not been found. The problem here is how to find it. Why should recombination of existing sequences, which share no homology with the wildtype, help at all in finding the wildtype? Mysteries of blind faith.

And have the authors, or anyone else, made new experiments that show how recombination can solve the limit they found? Not that I know. If you are aware of that, let me know.

Then you say:

Or their model of the fitness landscape might not be completely accurate.

Interesting strategy. So, if the conclusions of the authors, conclusions driven from facts and reasonable inferences, are not those that you would expect, you simply doubt that their model is accurate. Would you have had the same doubts, had they found that RM + NS could find easily the wildtype? Just wondering…

And again:

So between the limitations of their simulation of actual evolutionary processes and the limitations of the region of the landscape over which they gathered data, I don’t see how you can draw any particularly solid conclusions from that study.

Well, like you, I am not an expert of that kind of models. I accept the conclusions of the authors, because it seems that their methodology and reasonings are accurate. You doubt them. But should I remind you that they are mainstream authors, not certainly IDists, and that their conclusions must have surprised themselves first of all. I don’t know, but when serious researchers publish results that are not probably what they expected, and that are not what others expect, they must be serious people (except, of course, for the final note about recombination, but anyone can make mistakes after all! ).

Then your final point:

This seems to directly refute your claim that stepwise-beneficial mutations cannot produce functional proteins. They showed that it can.

No, for a lot of reasons:

a) We are in a scenario of tweaking an existing, damaged function to retrieve part of it. We are producing no new functional protein, just “repairing” as much as possible some important damage.

b) That’s why the finding of lower levels of function is rather easy: it is not complex at all, it is in the reach of the probabilistic resources of the system.

I will try to explain it better. Let’s say that you have a car, and that its body has been seriously damaged in a car accident. That’s our protein with its D2 domain replaced by a random sequence of AAs.

Now, you have not the money to buy the new parts that would bring back the old body in all its spendor (the wildtype).

So, you choose the only solution you can afford: you take a hammer, and start giving gross blows to the body, to reduce the most serious deformations, at least a little.

the blows you give need not be very precise or specific: if there is some part which is definitely too out of the line, a couple of gross blows will make it less prominent. And so on.

Of course, the final result is very far from the original: let’s say 2000 times less beautiful and functional.

However, it is better than what you started with.

IOWs, you are trying a low information fixing: a repair which is gross, but somewhat efficient.

And, of course, there are many possible gross forms that you can achieve by your hammer, and that have more or less the same degree of “improvement”.

On the contrary, there is only one form that satisfies the original request: the perfect parts of the original body.

So, a gross repair has low informational content. A perfect repair has very high informational content.

That’s what the rugged landscape paper tells us: the conclusion, derived form facts, is perfectly in line with ID theory. Simple function can be easily reached by some probabilistic resources, by RV + NS, provided that the scenario is one of tweaking an existing function, and not of generating a new complex one.

It’s the same scenario of malaria resistance, or of other microevolutionary events.

But the paper tells us something much more important: complex function, that with a high informational content, cannot be realistically achieved with those mechanisms, nor even in the most favorable NS scenario, with an existing function, and the opportunity to tweak it with high mutation rates and highly reproducing populations, and direct relevance of the function to reproduction.

Complex function cannot be found, not even in those conditions. The wildtype remains elusive, and, if the author’s model is correct, which I do believe, will remain elusive in any non design context.

And, if RV and NS cannot even do that, how can they hope to just start finding some new, complex, specific function, like the sequence of ATP synthase beta chain, or dynein, or whatever you like, starting not from an existing, damaged but working function, but just from scratch?

OK, this is it. I think I have answered your comments. It was some work, I must say, but you certainly deserved it!

Addendum:

By the way, in that paper we are dealing with a 139 AAs sequence (the D2 domain).

ATP synthase beta chain is 529 AAs long, and has 334 identities between E. coli and humans, for a total homology of 663 bits.

Cytoplasmic dynein 1 heavy chain 1 is 4646 AAs long, and has 2813 identities between fungi and humans, for a total homology of 5769 bits.

These are not the 16 – 28 bits of malaria resistance. Not at all.

OK, that’all for the moment. Again, I apologize for the length of it all!  🙂

351 Replies to “What are the limits of Natural Selection? An interesting open discussion with Gordon Davisson

  1. 1
    Origenes says:

    GPuccio,

    Surely there is enough here for several OP’s. Take for instance your elucidating take on the paper “Experimental Rugged Fitness Landscape in Protein Sequence Space” — ‘The rugged landscape experiment’ (part 9 & 10).

    I find it interesting to see that, as you point out, that the writers of the paper are not straightforward about the meaning of the number 10^70.

    From the paper (emphasis added):

    Based on the landscapes of these two different surfaces, it appears possible for adaptive walks with only random substitutions to climb with relative ease up to the middle region of the fitness landscape from any primordial or random sequence, whereas an enormous range of sequence diversity is required to climb further up the rugged surface above the middle region.

    “Enormous” is correct of course, but the unsuspected reader can easily think that there is an “enormous” amount of phages out there. But the reality is, that the number under discussion, 10^70, is, as you say, a probabilistic resource that is beyond the reach of nature. As a comparison, the sun contains 10^57 atoms of hydrogen.
    Later in the paper the reader is again required to read between the lines (emphasis added):

    By extrapolation, we estimated that adaptive walking requires a library size of 10^70 with 35 substitutions to reach comparable fitness. Such a huge search is impractical ….

    “Impractical in the lab?” some readers might think. No, here by impractical is meant “impractical in this universe” or “never going to happen”.

  2. 2
    gpuccio says:

    Origenes:

    Thank you for your first comment! After all, the existence of this OP is your merit. 🙂

    “Surely there is enough here for several OP’s.”

    Yes, I suppose there is a lot of stuff here. Maybe too much! 🙂

    The merit, or responsibility, for that is shared by me with Gordon Davisson, who raised so many good points with his posts that I had to win my inherent laziness and answer all of them.

    “I find it interesting to see that, as you point out, that the writers of the paper are not straightforward about the meaning of the number 10^70.”

    Yes, that’s a big number, isn’t it? Strangely, it’s more or less the number given a lot of time ago by Axe for the number of sequences needed to find a folding sequence, if I remember well. I am not saying that there is a relationship, but it’s a funny coincidence, isn’t it?

    I must say that I have loved the rugged landscape paper since the first time I read it. That 10^70 is so appealing, so much ID style!

    I must say that the paper was signaled to me by Petrushka (or was it Zachriel? No, I think it was Petrrushka) who strongly believed that it was hard evidence for the power of NS. As soon as I read it, I immediately thought that it was the opposite: absolute evidence of its limits.

    You are right about the use of words. I think that the authors realized all too well that they had some hot potato in their hands. So, they were honest and published the right facts and the right conclusions, because they obviously believe in their methodology and results, but I think they just tried to “smoothen” the impact by being smart with the words: “enormous range”, “impractical”, and, in the end, that completely unwarranted invocation of recombination as a convenient deus ex machina.

    However, I admire them and am really grateful to them for a very good research, one that really means a lot.

  3. 3
    Florabama says:

    I would be interested in how Haldane’s Dilemma plays into the question of beneficial mutation rates. It would seem to me that the cost of reproduction combined with “extremely rare” mutation rates, would make radical evolutionary change nearly impossible.

  4. 4
    gpuccio says:

    Florabama:

    I think you are right, Haldane’s considerations certainly add further difficulties to the “usefulness” of NS. The idea, if I understand it well, is that if many new beneficial traits have to be expanded simultaneously in a population, starting from different individuals and controlled by different genes, hey will compete for the expansion. I don’t know exactly how to add that factor to probability computations, I am not an expert in population genetics. I think, however, that NS is already defunct enough for the arguments that I have tried to detail, and Haldane’s problem can be the final blow. 🙂

    It is interesting that we often reason as though one single beneficial trait expanded at a time. Probably because that is what happens is some of the microevolutionary “models” from the lab or from highly specific situations, that are indeed the only models for NS and so come naturally to the mind of people when we speak of it.

    For example, consider simple antibiotic resistance, including malaria resistance, a situation where the environmental pressure (the antibiotic) is so strong that it can efficiently kill in a very short time all or almost all the individuals that have no feature of resistance, however small. In that context, resistance becomes naturally the only important beneficial trait, and it can expand undisturbed in the population, ad in a rather short time.

    The same is true for the rugged landscape experiment. For a phage, infectivity is survival, because phages cannot reproduce unless they infect the host cell. So here, again, loss of infectivity represents almost certain extinction, and any improvement in the damaged function, however small, becomes a passport for rapid expansion.

    But these are extreme cases, and somewhat artificial ones. In “normal” evolutionary history, a lot of new “traits” should be evolving at the same time, according to the theory. So, competition for expansion, Haldane’s problem, becomes a very real problem indeed.

    An important point that is often misunderstood is the importance of the expansion. The expansion of a mutation in the original population is critical to the theory. Indeed, each new variation, if it arises as a random event, arises in one individual, and will be confined to the descendants of that individual (let’s call it “the original clone”) unless it expands to great part of the population (or to all of it, if intermediates really must be cancelled in the process).

    It’s this expansion that can “lower” the probabilistic barriers that, as we well know, make the generation of complex functional information well beyond any universal threshold of impossibility.

    But the expansion can happen for two different mechanisms:

    a) Genetic drift. That happens to neutral or quasi neutral traits, that are expanded by this random mechanism. No reproductive avantage is needed here. The proble is that all mutations have the same probability to be expanded by genetic drift. therefore, genetic drift does not in any way lower the probabilistic barriers for anything. Itis absolutely neutral for our reasonings, therfore essentially irrelevant.

    b) Natural selection. Here the expansion is linked to a reproductive advantage. So, there is a necessity factor in action. As said, NS can potentially lower the probabilistic barriers in some very specific cases, and it certainly does that in known microevolutionary events. But it is a hugely limited process, which can only be invoked in very specific cases, and for the generation of few bits of functional information, not more than that.

    As I have tried to debate in my long OP! 🙂

    And you are right, Haldane’s dilemma certainly applies to restrict even more its role in natural evolutionary history.

  5. 5
    gpuccio says:

    EugeneS has posted about another aspect of NS in the old thread from which this OP is derived.

    I copy and answer his post here:

    GPuccio

    I am sorry if you have already addressed it above (I will need time to read it all at my pace). Could you elaborate on gene duplication a bit more in light of probabilistic barriers? In theory, gene duplication allows the neo-Darwinian model to traverse areas of the configuration space where function does not exist. The idea is that a duplicate (paralog) can change without being restrained by natural selection. As soon as it becomes functional, it is immediately subject to natural selection, but with a different function.

    I know this is too speculative, but could you say a bit more? People do mention gene duplication in discussing the capabilities of RV+NS.

    Thanks!

    OK, gene duplication.

    I would say that, while the role of gene duplication is certainly important in evolutionary history, there is great confusion about its relevance in a neo-darwinian theory.

    To simplify, I will describe two different scenarios:

    a) A gene is duplicated, remains functional, and the functional copy undergoes RV + NS so that a new gene is obtained.

    b) A gene is duplicated and inactivated. IOWs, it becomes a pseudogene. Then it undergoes the process of RV + NS.

    The two situations are completely different. While there may be intermediate scenarios, I think that those two can clarify well the possible role, or non role, of NS.

    I will say immediately that b) is probably the important issue. However, let’s say something about a).

    Eugene, you mention in your post an aspect of NS that I have not really covered in my OP: that, indeed, NS can and does work, in many cases, against neo darwinian evolution. As you say, it can restrain evolution.

    The reason is simple: NS is of two kinds.

    1) Negative NS is the strongest form, the form that is easily recognizable in mature. It is the selection against variation that reduces the function of an existing protein. Another techinical term for it is “purifying selection”.

    Negative selection is a strong, universal force. It is the force that keeps the functional sequence of proteins rather constant, operating against deleterious mutations.

    So, this powerful force has the effect of keeping the existing functional information as it is. It can only tolerate neutral or quasi-neutral variation.

    Of course, if a functional gene is supposed to change into something different, either in its original form or in a duplicate functional form, negative selection will act against that change. So, it is in general a force against neo-darwiniam evolution.

    2) The aspect of NS which should act in favor of it is positive selection: the fixation, by expansion to the whole population or a significant part of it, of a beneficial mutation which confers a reproductive advantage.

    Now, this type of mechanism is certainly much rarer than negative selection: indeed, it is very difficult to document it in most cases, even if there are clear cases of positive selection in action, like the cases of microevolution we have discussed in the OP.

    I think that the only reasonable scenario where a gene duplication could perhaps generate a new functional gene by the neo-darwinian mechanism is the following:

    A gene is duplicated, and remains functional. While the original gene ensures that the old function is satisfied, the new gene undergoes small variations, 1 – 5 AAs, at the active site, and is transformed into a similar gene, with more or less different biochemical activity.

    This could be a mechanism that generates diversification in an existing protein family, for example.

    The point is: most of the sequence and structure of the old gene, here, will be conserved. That’s why it is a good thing that the duplicated gene remains functional, so that negative selection can preserve that bulk of sequence, structure and functionality.

    On the other hand, while the folding and the general structure of the protein remain the same (IOWs, we remain in the original island of the original protein family) the active site can undergo some small variation that changes its biochemical affinity for substrates, and so in the end provides a different range of activity and function.

    This variation at the active site is usually in the microevolutionary range (as I said, 1 – 5 AAs), so it could be potentially in the range of very good biological probabilistic resources.

    So, what do I believe about this scenario? I believe that those cases are borderline: they could be extreme cases of neo-darwinian microevolution, or very simple cases of designed macroevolution.

    Therefore, it is wise IMO not to focus a design inference on that type of processes: we have indeed a lot of scenarios where the informational jump is hundreds of times bigger, beyond any possible reach of the neo-darwinian theory.

    For example, all cases of appearance of a new protein superfamily, or in general of a huge quantity of new information in a protein. You can take a look at my OPs about the informational jump in vertebrate proteome to find a lot of such examples.

    OK, I will discuss the b) scenario in next post.

  6. 6
    gpuccio says:

    EugeneS:

    The most interesting case is, IMO, when a gene duplication is followed by the inactivation of the gene, which becomes a pseudogene.

    As you say:

    “The idea is that a duplicate (paralog) can change without being restrained by natural selection.”

    Indeed, an inactivated gene isn’t subject any more to the restraining effects of negative selection: it is free to change in all possible ways.

    On the other hand, until it remains inactivated, it isn’t any more subject to the possible expansion of positive selection, in case some beneficial mutation should happen.

    The point is simple: a pseudogene should behave exactly like any other non coding DNA sequence that is not functional. It is not subject to any form of NS, and its expansion can only happen because of neutral genetic drift.

    You say:

    “In theory, gene duplication allows the neo-Darwinian model to traverse areas of the configuration space where function does not exist.”

    But the simple fact is that those areas cannot be traverses, because the whole probability barriers that we well know act against that possibility.

    As soon as a gene is inactivated, it is fully subject to neutral variation. In the absence of any negative selection, it will soon lose any relation to the original functional sequence of the original gene. After some time, it can be considered at the same level of any random nucleotide sequence.

    And that must happen, if the purpose is to reach another functional island, a new functional gene that is completely unrelated, at sequence level, to the original one. For example, a new superfamily.

    But the probabilities that a random sequence finds a new functional island are practically zero. Do not be confused by the rugged landscape paper: there, we have an existing, damaged function. No new function is found, only some fixing of the damage is attained, as I have argued in the OP.

    To get to functional sequences from random sequences by random variation means to find functional information by chance alone. Even darwinists recognize that it is impossible.

    Moreover, everyone in the darwinian field seems to ignore the problem in this other statement:

    ” As soon as it becomes functional, it is immediately subject to natural selection, but with a different function.”

    As soon as???

    But we are discussing of a non coding gene. OK, it changes. Maybe its new sequence could be vaguely functional. Let’s ignore for the moment that it needs a promoter, one or more enhancers, a regulation system, an integration with what already exists, to be really “functional”, least of all naturally selectable.

    But let’s be serious, that sequence is not transcribed and translated, Even if it transcribed, it is certainly not translated. How can the living being know that it has become functional?

    I think that neo-darwinists imagine a cell, or organism, where all non coding sequences are constantly transcribed and translated, filling the cell with junk of all kinds, so that the rare peptide that, as soon as it is translated, really helps may be expanded by NS!

    Or a scenario where non coding sequences not only, by magic, acquire a configuration that can correspond, by symbolic translation to a functional peptide sequence, and at the right moment, but not before, acquire a starting sequence and become translated ORFs, ready with their promoter, enhancers, TFs, and so on, to be of immediate relevant help to the cell and be expanded and fixed, before any new neutral variation can change the precious result.

    OK, I think this is folly, utter folly.

    I do believe that new genes come from non coding DNA, be it a pseudogene or any other sequence, but that process is a design process. The gene is prepared so that, once activated, it will be a functional protein. Then, and only then maybe NS can be of some collateral help.

    And, in many cases, the configuration of the future gene at the level of non coding DNA, is realized by transposon activity. We have some evidence of that.

    Transposons are the most likely tool of biological design.

  7. 7

    Excellent post and comments. Thank you all.

  8. 8
    Origenes says:

    GPuccio @2

    Strangely, it’s more or less the number given a lot of time ago by Axe for the number of sequences needed to find a folding sequence, if I remember well.

    Were you perhaps alluding to the following?

    The prevalence of low-level function in four such experiments indicates that roughly one in 10^64 signature-consistent sequences forms a working domain. Combined with the estimated prevalence of plausible hydropathic patterns (for any fold) and of relevant folds for particular functions, this implies the overall prevalence of sequences performing a specific function by any domain-sized fold may be as low as 1 in 10^77, adding to the body of evidence that functional folds require highly extraordinary sequences.

    [Axe, Gauger ‘Estimating the Prevalence of Protein Sequences Adopting Functional Enzyme Folds’, link]

    GPuccio: I must say that the paper was signaled to me by Petrushka (or was it Zachriel? No, I think it was Petrrushka) who strongly believed that it was hard evidence for the power of NS. As soon as I read it, I immediately thought that it was the opposite: absolute evidence of its limits.

    Petrushska must blame the authors of the paper for his confusion. As things are presented these days, one has to be an expert to understand what is not said in sections as ‘introduction’ and ‘discussion’.

  9. 9
    Florabama says:

    Thank you, gpuccio. Fascinating OP and responses.

  10. 10
    gpuccio says:

    Origenes:

    Yes, that’s the paper I meant. Thank you.

    Petrushka has always been rather stubborn, but in general he believed in what he said, which is not always true of all our kind interlocutors.

  11. 11
    gpuccio says:

    Florabama:

    Thank you! 🙂

  12. 12
    Gordon Davisson says:

    Hi, gpuccio; thanks for your replies! I’m going to try to work through your discussion, but I think I should set a few expectations.

    First off, I am neither a prompt nor a reliable correspondent. Basically, I find that the actual process of writing slow and a bit painful; on the other hand, I’m really good at procrastinating! So I tend to mull over what I should say for quite a long time before I manage to convince myself to actually start typing in a reply. If I ever do convince myself to start…

    (& then I post and inevitably find a half-dozen mistakes that I should have noticed while I was profreading, and frantically try to fix them within the edit window without adding any new mistakes in the process…)

    Anyway, I’ll try to be at least semi-reliable about getting back to you, but I’ll almost certainly not be at all prompt. Sorry about that. On the other hand, I do think a well-mulled-over discussion is generally better quality than something quickly dashed off.

    Second, I’m not going to try to address all of the points you’ve raised. There’s a tendency in these discussions (and already in this one) for people to start arguing about topic A, then find that they also disagree about related topics B, C, and D, and start arguing about those, which leads to E through Z… and by the time everyone’s given up on the discussion, we’re well into the Greek alphabet, nothing ever got settled, and all anyone learned is that everyone on the other side is wrong about everything.

    To avoid the problem of having to settle everything in order to settle anything, I’d like to try to keep the subject of discussion relatively contained. So I’ll try to avoid going after every interesting digression that comes along (and try to avoid raising too many side topics myself). But that means we’d better agree, before getting too far, on what we’re actually trying to talk about. I think the central topic of discussion is your contention that:

    * Evolution cannot produce new complex, functional systems (e.g. proteins) without being led to them via step-by-step-beneficial paths.
    * Such paths don’t exist.

    Is that a fair summary of your claim? And if so, would you agree that this should be the central topic here?

    BTW, having said this the first two specific topics I plan to address aren’t actually directly related to that. But they’re foundational issues that IMO we have to come to some sort of agreement about, before we have any real chance of meaningful discussion of that central issue. Specifically, I think we need to at-least-mostly agree on what evolution is (i.e. whether RM + NS is the “official model”), and on how to evaluate evidence as supporting one or another view (i.e the similarity between Human orphan proteins and chimp non-coding sequnces).

    Ok, last item for the moment: I need to own up to a mistake. Concerning the rugged landscape paper, you said:

    1) The whole protein is implied in infectivity

    2) Only the central domains has been replaced by random sequences

    I missed this in my read through the paper, and was thinking that they’t started with an entirely random polypeptide. This does significantly weaken some of the conclusions that I drew from the paper (although I think some also remain intact).

    (AIUI there are also some other studies that did start from at or near entirely random sequnces and evolve function, but I’d have to do a fair bit of research before I’d be familiar enough with them to argue from them. So I’ll put that off for later maybe.)

  13. 13
    gpuccio says:

    Gordon Davisson:

    Thank you very much for coming back to the discussion! 🙂

    I really appreciate what you say in this last post. I agree with all.

    The purpose of moving our discussion here was exactly to give it some more proper space, to invite everyone interested to join, and to give you a chance, as you wished, to give further contributions with your pace.

    As I already said, I am very grateful to you for the way you raised so many relevant points, allowing me to express some ideas about them. There is nothing better than a good interlocutor in intellectual confrontation.

    * Evolution cannot produce new complex, functional systems (e.g. proteins) without being led to them via step-by-step-beneficial paths.
    * Such paths don’t exist.

    Is that a fair summary of your claim? And if so, would you agree that this should be the central topic here?

    Absolutely! That is the core of all the discussion, and of ID theory itself, IMO.

    Specifically, I think we need to at-least-mostly agree on what evolution is (i.e. whether RM + NS is the “official model”), and on how to evaluate evidence as supporting one or another view (i.e the similarity between Human orphan proteins and chimp non-coding sequnces).

    As you said very well, for me the problem is not “evolution”, but the explanation of how it happens, IOWs of how complex functional information is generated in biological beings.

    Theregore, the debate is, IMO, between:

    a) A design explanation, requiring definite intervention from some conscious intelligent agent to input new complex functional information

    b) Any other non design explanation that has any potential explanatory power

    I usually debate neo darwinism, in particular the RV + NS algorithm, because in essence I am convinced that there is nothing else in the non design field that really deserves debate.

    But, of course, I am ready to discuss anything else that is suggested, by you or others.

    It could be useful to remind (but I am sure that you are aware of that) that I fully accept common descent, ad that my idea of ID is a model of guided common descent. So, there is no reason to debate common descent, because I suppose we agree on that point.

    Again, the only problem is how the new complex functional information comes into existence.

    So, you are really welcome to contribute as you like, and with the pace that you like. Anything will be deeply appreciated. I really love your intellectual honesty! 🙂

  14. 14
    tribune7 says:

    Great points gpuccio

    –Note that all of these would be considered beneficial mutations:
    * Minor changes to an existing functional thing (protein, regulatory region, etc) that improve its function slightly.
    * Minor changes to an existing functional thing that change its function slightly, in a way that makes it fit the organism’s current environment better.
    * Changes that decrease function of something that’s overdoing its role (e.g. the mutation discussed here, which winds up giving people unusually strong bones).–

    Few deny this and it is not controversial. Someone defending neo-Darwinianism as an axiomatic explanation for all biodiversity should not even bring them up as it just muddies the water.

  15. 15
    tribune7 says:

    –Evolution cannot produce new complex, functional systems (e.g. proteins) without being led to them via step-by-step-beneficial paths.–

    Wouldn’t a better way of saying this be “Evolution has not been shown to be able to produce new complex, functional systems (e.g. proteins) without being led to them via step-by-step-beneficial paths” ?

    Saying “cannot” is claiming to prove a negative. It’s the one who says “can” or “did” from whom it is reasonable to demand an explanation.

  16. 16
    ET says:

    Natural selection is an eliminative process and you don’t get complex adaptations by merely eliminating the less fit.

  17. 17
    Dionisio says:

    I like the seriousness of this discussion between GP and GD.
    Looking forward to seeing what transpires from it.

  18. 18
    Corey Delvine says:

    “Natural selection is an eliminative process and you don’t get complex adaptations by merely eliminating the less fit.”

    Isn’t evolution more complex than that?

    Isn’t evolution dependent on constant small changes which are then selected for based on their effects in the organism or between the organism and it’s environment?

  19. 19
    gpuccio says:

    Corey Delvine:

    “Isn’t evolution more complex than that?

    Isn’t evolution dependent on constant small changes which are then selected for based on their effects in the organism or between the organism and it’s environment?”

    Evolution according to the neo-darwinian algorithm is based on two different processes: random variation, which is probabilistic, and natural selection, which is in a way a necessity process.

    The main topic in this OP and in the comments is exactly NS: how it works, and what are its limits.

    Have you read the OP?

    ET’s statement, to which you objected, was the following:

    “Natural selection is an eliminative process and you don’t get complex adaptations by merely eliminating the less fit.”

    Now, there can be no doubt that NS is an eliminative process. Indeed, there are two kinds of NS, as I have discussed in my comment #5 here. I paste here the relevant part:

    The reason is simple: NS is of two kinds.

    1) Negative NS is the strongest form, the form that is easily recognizable in mature. It is the selection against variation that reduces the function of an existing protein. Another technical term for it is “purifying selection”.

    Negative selection is a strong, universal force. It is the force that keeps the functional sequence of proteins rather constant, operating against deleterious mutations.

    So, this powerful force has the effect of keeping the existing functional information as it is. It can only tolerate neutral or quasi-neutral variation.

    Of course, if a functional gene is supposed to change into something different, either in its original form or in a duplicate functional form, negative selection will act against that change. So, it is in general a force against neo-darwiniam evolution.

    2) The aspect of NS which should act in favor of it is positive selection: the fixation, by expansion to the whole population or a significant part of it, of a beneficial mutation which confers a reproductive advantage.

    Now, this type of mechanism is certainly much rarer than negative selection: indeed, it is very difficult to document it in most cases, even if there are clear cases of positive selection in action, like the cases of microevolution we have discussed in the OP.

    Now, there can be no doubt that negative selection is eliminative.

    But positive selection is eliminative too, in a sense, because it eliminates the previous form of the allele, which has become less fit, allowing the new form to expand.

    So, I think that the first part of ET’s statement is certainly true.

    The simple truth is that, according to the neo-darwinian algorithm, the only real engine that generates new functional information is random variation (RV).

    That’s why ET states, in the second part of his comment, that “you don’t get complex adaptations by merely eliminating the less fit.”.

    And he is right!

    Practically everyone agrees that RV has not the probabilistic power to generate even a tiny part of the huge complex functional information we observe in biological beings.

    Of course, neo-darwinists believe that NS, by the process of expansion of simpler beneficial mutations, can help in the process, and lower the probabilistic barriers that stand against RV as an engine of complex novelty.

    I am absolutely convinced that that idea is not true. And my long OP here is almost completely dedicated to explaining why I believe that.

    So, if you have the time, please read it. And, if you want, please comment on it.

    However, welcome to the discussion.

  20. 20
    Mung says:

    gpuccio:

    Evolution according to the neo-darwinian algorithm is base on two different processes: random variation, which is probabilistic, and natural selection, which is in a way a necessity process.

    I disagree with this. It is natural selection that is probabilistic. Random variation is just magic.

  21. 21
    Corey Delvine says:

    “Natural selection is a force against evolution”
    Interesting, because I’d wager that every evolutionary biology professor will at some point during the semester say something along the lines of:
    “natural selection is the main driving force behind evolution”
    Now why would your understanding of natural selection and an evolutionary biologist’s understanding be complete opposites?
    …Hmm I wonder…..

    You break natural selection into two forms, but these two forms are actually one in the same.
    Immediate removal (negative selection) of deleterious changes, while allowing neutral (or nearly neutral) and potentially beneficial changes to remain is what, over time, enriches the gene pool for beneficial changes (positive selection).
    You seem to have (subconsciously?) realized this because immediately after you copy/pasted from your previous post, you go on to say that “positive selection is eliminative too.”
    Yes it is, because negative and positive selection are two parts of a whole.

    ET says “you don’t get complex adaptations by merely eliminating the less fit” and you think he/she is referring to the fact that random variation is required as well.
    I read what ET said and it seems to me that he/she is trying to refute evolution by redefining it as only “eliminating the less fit” and then saying “that’s not enough.”
    (Is this what’s called a strawman argument?)
    I assumed this largely because it’s ET’s first comment on this post and it wasn’t directed to any one in particular. Either way, that is how I interpreted it and others could do the same just as easily.

    Now, I’m not sure exactly who “neo-darwinists” are, but
    “natural selection, by the process of expansion of simpler beneficial mutations, can help in the process, and lower the probabilistic barriers that stand against RV as an engine of complex novelty”
    is a good explanation of the process of evolution, in my opinion.
    I think that you disagree with it because you severely underestimate the types and complexities of “random variation”.

    Random variation is NOT just point mutations, which you probably already know.
    But do you know the variety of changes that are actually encompassed by “random variation”?

    It is quite astounding.
    And even relatively simple changes can have huge effects on organisms at both the cellular and the organismal levels.
    (Both potentially positive and negative effects I might add.)

  22. 22
    Dionisio says:

    gpuccio @19:

    “Of course, neo-darwinists believe that NS, by the process of expansion of simpler beneficial mutations, can help in the process, and lower the probabilistic barriers that stand against RV as an engine of complex novelty.”

    Well, why not?

    There are known examples of fast macroevolution: a pumpkin converted into an elegant carriage. Mice turned into beautiful horses. That story has been around quite long. Who needs more proof?
    🙂

  23. 23
    gpuccio says:

    Corey Delvine:

    You say, supposedly quoting me:

    “Natural selection is a force against evolution”

    This is definitely quote mining. What I said is:

    Of course, if a functional gene is supposed to change into something different, either in its original form or in a duplicate functional form, negative selection will act against that change. So, it is in general a force against neo-darwiniam evolution.

    Emphasis added for your convenience.

    The idea is simple enough: negative, or purifying, selection just eliminates changes, rso it preserves what already exists. It is rather obvious that it is a force acting against change, therefore against evolution. At best, it is a force which preserves what has been already created (whatever the mechanism that created it).

    When your imaginary professore says:

    “natural selection is the main driving force behind evolution”

    he is obviously referring to positive selection, not to purifying selection. So, I seem to be in agreement, for what it is possible, with your imaginary professor, but not with your more or less purposeful misunderstanding of what I said.

    You say:

    “You break natural selection into two forms, but these two forms are actually one in the same.
    Immediate removal (negative selection) of deleterious changes, while allowing neutral (or nearly neutral) and potentially beneficial changes to remain is what, over time, enriches the gene pool for beneficial changes (positive selection).
    You seem to have (subconsciously?) realized this because immediately after you copy/pasted from your previous post, you go on to say that “positive selection is eliminative too.”
    Yes it is, because negative and positive selection are two parts of a whole.”

    But I agree with that! Of course they are two parts of the whole. But each part has different roles, and does different things.

    So, there is nothing subconscious in my argument. I am well aware that both processes are eliminative, and based on the concept of reproductive advantage, but you seem to forget that what is eliminated is different in the two cases:

    a) In negative selection, it is any new mutation which is deleterious enough which is eliminated

    b) In positive selection, it is the old population with the old form of the allele which is eliminated

    So, while both processes are based on eliminating, the two processes are completely different in their result and their roles.

    Frankly, your attempt at denying the differences between negative selection and positive selection, which are obvious in all the evolutionary science, is rather pitiful.

    For example, you may know that when we compute a Ka/Ks ratio, negative and positive selection can be easily distinguished:

    From Wikipedia (page: Ka/Ks ratio):

    The Ka/Ks ratio is used to infer the direction and magnitude of natural selection acting on protein coding genes. A ratio greater than 1 implies positive or Darwinian selection (driving change); less than 1 implies purifying or stabilizing selection (acting against change); and a ratio of exactly 1 indicates neutral (i.e. no) selection.

    Emphasis mine.

    Anyone can interpret what ET said as he likes. My point is not to interpret others (who can certainly speak for themselves) but to clarify things. Even if I start from what someone says, the things I say are my responsibility, and none else’s.

    You say:

    “Now, I’m not sure exactly who “neo-darwinists” are, but
    “natural selection, by the process of expansion of simpler beneficial mutations, can help in the process, and lower the probabilistic barriers that stand against RV as an engine of complex novelty”
    is a good explanation of the process of evolution, in my opinion.”

    I am happy that you agree with my description of the process.

    I call “neo-darwinists” all those who accept the neo-darwinian paradigm, IOWs the “Modern Synthesis”, and all its following forms, based on RV + NS as an explanation for biological complexity.

    “I think that you disagree with it because you severely underestimate the types and complexities of “random variation”.”

    That’s absolutely not true. I don’t underestimate random variation at all. I have a very clear idea of what it is, and of what it can do.

    “Random variation is NOT just point mutations, which you probably already know.”

    Yes. As you imagined, I already know that. You may have noticed that I use the form: RV + NS for the neo-darwinian algorithm, and not RM + NS. That includes all forms of random variation, certainly not only SNVs.

    You say:

    “But do you know the variety of changes that are actually encompassed by “random variation”?”

    Yes, I do.

    “It is quite astounding.”

    Why? You just apply any form of RV to a computer code, and you will get similar “astounding” results. I am not astounded at all!

    “And even relatively simple changes can have huge effects on organisms at both the cellular and the organismal levels.”

    Of course. Even one bit difference can crash a whole system.

    “(Both potentially positive and negative effects I might add.)”

    Both minimally positive effects, and huge negative aspects, I would add. Please, read my OP.

    Look, I will try to make a serious argument about that, if you want to listen.

    Random variation can be of many different kinds: single nucleotide variants, both synonimous and non synonimous, or in non coding DNA, indels, translocations, duplications, exon shuffling, whatever you like.

    But, when we reason about probabilities in a random walk (which is the only appropriate way to deal quantitatively with RV in biology), then each single event of RV, whatever it is, gives a result which has always the same meaning: a new state.

    With each event of RV we are reaching a new state. One state among all the possible states.

    The probability of each possible state to be reached is similar, because the event is by definition random.

    Of course, there are exceptions, mainly the fact that SNVs, being by definition small variations, usually can reach only sequences in the immediate landscape, and not others. But there is an important exception even to that: frameshift mutations. As you certainly well know.

    So, what is all this excitement about the “variety of changes that are actually encompassed by “random variation”” about?

    One change = one new state. What is astounding in that?

    If you have to reach the 300+ conserved aminoacids in ATP synthase beta chain which are necessary for its function, how is the “variety of changes” going to help you?

    In no way. A specific sequence of 300 aminoacids is one state among about 10^390.

    Every time you have a RV event, you are “testing” one of those 10^390 states, whatever the nature of the variation.

    This is the simple truth.

  24. 24
    ET says:

    Corey:

    Isn’t evolution more complex than that?

    Natural selection is an eliminative process that eliminates the less fit over time. See Mayr “What Evolution Is”.

    Isn’t evolution dependent on constant small changes which are then selected for based on their effects in the organism or between the organism and it’s environment?

    Nature doesn’t select for anything. Whatever is good enough gets to survive and have a chance at reproducing.

  25. 25
    ET says:

    Corey:

    ET says “you don’t get complex adaptations by merely eliminating the less fit” and you think he/she is referring to the fact that random variation is required as well.

    Natural selection includes random variation. Random variation is the first step of natural selection- Mayr “What Evolution Is”

    I read what ET said and it seems to me that he/she is trying to refute evolution by redefining it as only “eliminating the less fit” and then saying “that’s not enough.”

    What? I am not trying to refute evolution. I was just making a point about natural selection, the alleged designer mimic which is clearly nothing of the sort.

    Random variation is NOT just point mutations, which you probably already know.

    Actually that is part of the debate as ID says not all mutations are random. The only mutations you can justifiably say are random are point mutations. See Spetner, “Not By Chance” 1997.

    he/she

    I am in a state of flux.

  26. 26
    Corey Delvine says:

    So let me get this straight.
    You claim that “negative selection works against evolution.”
    Then agree that positive selection is the driving force behind evolution.
    Then you go onto agree that “negative and positive selection are two parts of a whole.”
    So which is it?

    “The two processes are completely different in their result and their roles”
    If we look only at amino acid sequence, sure.
    But don’t miss the forest for the trees.
    Looking at the bigger picture, both processes are part of the foundation of natural selection and both shift a population toward greater fitness in a given environment.

    Anyways, you’ve only mentioned a small number of the potential amounts of variations.
    You are missing many of the interesting ones, and just sweeping them all under the rug like that does not look good on your part.

    Are you aware that deletion of a single HOX gene has been shown to shift the rib cage up and down the spine of model organisms?
    How about the fact that a chromosomal inversion in fruit flies drives growth of legs where antennae would normally be?
    While I am mentioning these more for shock value rather than actual biological function, these massive changes at the organismal level are driven by small or even no change in DNA sequence.
    Are you sure that you understand the breadth of what “random variation” entails?

    You throw around huge numbers like 10^390, but it only affects people who don’t understand how evolution works.
    We did not walk our way to ATP synthase one amino acid at a time.
    No biologist would ever claim that’s how it happened.

  27. 27
    ET says:

    Corey:

    We did not walk our way to ATP synthase one amino acid at a time.

    True. There is no doubt that ATP synthase was Intelligently Designed. Just look at its structure-> 2 different subunits held just the right distance apart by an external docking mechanism? And nature just happened to figure out that the way to get ATP is by squeezing captured P’s onto captured ADP’s? Squeeze, mind you and not chemically catalyzed.

    Are you aware that deletion of a single HOX gene has been shown to shift the rib cage up and down the spine of model organisms?

    So you want us to believe that developmental control genes just happened to come along, work and because of that were kept? Are you are that the presence of developmental controllers and regulation are evidence for ID? Or do you have any evidence or way to test the claim that non-telic processes did it?

  28. 28
    Dionisio says:

    The answer to the following question is given in the OP, but let’s clarify it again:

    When the term ‘evolution’ is used in this discussion, does it refer to microevolution, like the famous Galapagos finch case or to macroevolution, like the often comparison between chimps and humans, or both, or either one, or none of those, or something else?

    For example, the below links refer to the Galapagos finch case:

    https://uncommondescent.com/intelligent-design/epigenetics-may-explain-how-darwins-finches-respond-to-environment/

    https://uncommondescent.com/evolution/researchers-darwins-finches-not-typical-example-of-evolution-at-all/

  29. 29
    Corey Delvine says:

    “Squeeze, mind you and not chemically catalyzed.”

    Wow.

    Why on Earth did I think it was a good idea to talk to you in the first place.

    If you get anything out of this, ET, please let it be the fact that how a protein originally evolved and how proteins have evolved from other proteins are two very different questions.

    Something you guys like to forget.

  30. 30
    ET says:

    Corey, I don’t understand your issue. Do you not understand how ATP is made? Here is a clue- How Cells Make ATP: ATP Synthase:

    The first state is called the “open” state. In the “open” state ADP and Pi can bind to the active site. In the second state, called the “loose” state, ADP and Pi are locked in place and cannot be released. In the third state (the “tight” state”) ATP is formed as ADP and Pi are squeezed together.

    This mechanism of ATP synthesis is called the binding change mechanism.

    You’re welcome for the lesson.

    If you get anything out of this, ET, please let it be the fact that how a protein originally evolved and how proteins have evolved from other proteins are two very different questions.

    No one said nor implied otherwise. You have a strawman issue also.

  31. 31
    Corey Delvine says:

    “Squeeze, mind you and not chemically catalyzed.”

    First, bringing two molecules near each other and in the right orientation to catalyze a reaction is not specific to ATP synthase.

    And second, if you think there is no chemistry involved in generating ATP….well

    …don’t let the door hit you on the way out.

  32. 32
    gpuccio says:

    Corey Delvine:

    What’s the problem with you? Are you really trying to understand what I say? It doesn’t seem so.

    You say:

    So let me get this straight.
    You claim that “negative selection works against evolution.”
    Then agree that positive selection is the driving force behind evolution.
    Then you go onto agree that “negative and positive selection are two parts of a whole.”
    So which is it?

    I thought I had been clear enough. But let’s try again.

    a) I claim that “negative selection works against evolution.” Wikipedia seems to agree. Biologists seem to agree.

    You don’t agree. OK, you are entitled to your own opinion. Perhaps you could even try to give some reasonable argument if why negative selection act in favor of evolution, or is netral, or whatever you believe. You know, it’s called discussion.

    Until then, I will stick with my reason, with Wikipedia, and with biologists.

    b) I certainly agree that positive selection is the only driving force behind darwinian evolution. because it is true.

    Of course, I don’t agree that such a driving force can drive anything more than those limited cases of microevolution that I have debated in great detail in my OP, or similar cases.

    As should be evident to whever has read even a small part of what I have written in this thread, I certainly don’t believe that such a driving force can generate new complex functional information. Indeed, I am absolutely certain of the opposite. And I try to explain why in everything I say here.

    c) I certainly agree that “negative and positive selection are two parts of a whole.” But I have also cleraly specified that they are anyway responsible for two very different processes, and that, while they are coth eliminative processes, the whole difference is given by what is eliminated.
    I thought that was clear. Where is your problem? Is my English so bad?

    So, why ever do you ask:

    So which is it?

    It’s very simple. It’s a + b + c. There is no contradiction at all.

    That said, you think whatever you lik.

    “The two processes are completely different in their result and their roles”
    If we look only at amino acid sequence, sure.

    Ehm, what should we look at? I must really be dumb. I thought that biological information was written, at least for the part we understand better, in protein coding genes, and that the only role of protein coding gene was to determine the aminoacid sequence of proteins, from which derive the secondary and tertiary structure and therefore the function.

    But now, in your wisdom, you have clarified to me that I must not look only to the aminoacid sequence. I must also look at… what exactly? I must have missed something in your discourse… and I will see the light.

    But don’t miss the forest for the trees.
    Looking at the bigger picture, both processes are part of the foundation of natural selection and both shift a population toward greater fitness in a given environment.

    Ah, that’s the trick… I must look at the forest. Very poetical!

    But, I thought that the forest was made of trees. You know, if you have no trees, but for example only blades of grass, what you get is a meadow, not a forest. However you look.

    Anyways, you’ve only mentioned a small number of the potential amounts of variations.
    You are missing many of the interesting ones, and just sweeping them all under the rug like that does not look good on your part.

    Let’s see what I missed…

    Are you aware that deletion of a single HOX gene has been shown to shift the rib cage up and down the spine of model organisms?
    How about the fact that a chromosomal inversion in fruit flies drives growth of legs where antennae would normally be?
    While I am mentioning these more for shock value rather than actual biological function, these massive changes at the organismal level are driven by small or even no change in DNA sequence.
    Are you sure that you understand the breadth of what “random variation” entails?

    Ah, this is almost an argument. My compliments!

    OK, did you miss the part where I wrote, in anser to you:

    “Even one bit difference can crash a whole system.”

    and

    “Both minimally positive effects, and huge negative aspects, I would add.”

    I agree with you with the “huge negative aspects”. And you are giving good examples of them.

    “deletion of a single HOX gene has been shown to shift the rib cage up and down the spine of model organisms”

    Wonderful! Really a huge functional gain.

    “a chromosomal inversion in fruit flies drives growth of legs where antennae would normally be”

    which is, I suppose, a welcome novelty.

    OK, show me how a simple change in the genetic code can generate a new complex proteins, a completely new functional sequence, or change limbs into functional wings. That would be interesting.

    Hox genes are final effectors in determining the location of parts. Of course, if you derange them, you derange the final result.

    Should I be astounded? Why? Try to take out a single transistor in some key position of some very complex circuit, and see what you can do.

    Should we be amazed at the wonderful power of electronic damage?

    You throw around huge numbers like 10^390, but it only affects people who don’t understand how evolution works.

    (emphasis mine)

    Why am I not surprised to hear that? 🙂

    (Dionisio, please, don’t laugh out so loud! 🙂 )

    We did not walk our way to ATP synthase one amino acid at a time.

    I am sure of that. But I supposed that was the standard neo-darwinian idea. OK, maybe two aminoacids at a time. Maybe three.. no, three is too difficult.

    No biologist would ever claim that’s how it happened.

    I don’t know. I have never heard any biologists, let’s say any interlocutor at all in the neo-darwinist field, claim anything about how it happened. And I have asked that question about ATP synthase beta chain for years, here.

    You know? You could be the first.

    OK, please, try!

  33. 33
    gpuccio says:

    Corey Delvine:

    Just read this from you:

    how a protein originally evolved and how proteins have evolved from other proteins are two very different questions

    Maybe. So, you certainly have two very different answers!

    Don’t be shy, give them to us.

  34. 34
    ET says:

    Corey:

    First, bringing two molecules near each other and in the right orientation to catalyze a reaction is not specific to ATP synthase.

    Force of pressure

    And second, if you think there is no chemistry involved in generating ATP….well

    I never said nor implied such a thing.

    Look when a polypeptide is formed it is formed via a catalyst that speeds up a chemical reaction. ATP synthase uses force of pressure to squeeze another atom onto an existing molecule. It is a physical rather than chemical force that creates ATP

  35. 35
    Gordon Davisson says:

    I see the discussion is turning into a melee (sigh). I’m going to basically ignore everything since gpuccio @ 13, and follow my original plan of discussing whether whether RM + NS is the official model of evolutionary theory (and then segue back to talking about the “rugged fitness landscape” paper). The particular comment you made that’s set me off is:

    Well, they [the authors of the paper] didn’t use “a simplified model of evolution”. They tested the official model: RM + NS. And it failed!

    …which I have to strongly disagree with. First, because there’s no such thing as an official model of evolution, and second because if there were it wouldn’t be just RM + NS.

    Trying to define an “official model” of evolution is a lot like trying to define one for genetics or chemistry or geology. All of these study complex, diverse, and messy categories of phenomena; they’re fields that are constantly in flux, with new things being learned and sometimes things we thought we knew being reassessed. Trying to nail down exactly what the state of knowledge of any of these fields at any given time would be an exercise in nailing jello to a tree. What you can say, at any given time, is that there’s general agreement about some core knowledge about the phenomena, surrounded (metaphorically) by a cloud of reasonably-well-supported hypotheses, which are surrounded in turn by many many open and/or controversial questions. But the question of how settled various questions are is itself subjective, so you can’t even objectively define how solid various parts of our understanding are.

    This isn’t anything special about evolution; it’s just how things are in any dynamic field of science.

    Now, despite the difficulty of defining exactly what the official model/settled core/whatever of evolution is, I can confidently say that RM + NS is not it. At the very least, it needs to include genetic drift as a core process. But RM + NS + GD would still be a drastic oversimplification.

    Really, it’d need to consist of a catalog of all of the various evolutionary mechanisms we’ve discovered over the decades. Most of them can at-least-sort-of be categorised as types of mutation (e.g. unequal crossing over, transposon and retroviral insertion), selection (biased gene conversion, meiotic drive), or genetic drift (the founder effect, hitchhiking), but not always, and not always into just one category. The hypermutable state model, for example, involves an unusual type of mutation tightly coupled to an unusual type of selection, so it sort of fits into both of those categories. On the other hand, endosymbiosis and homologous recombination don’t really fit into any of the categories.

    Any model of evolution that doesn’t include all of the above mechanisms (and others I didn’t list / don’t even know about) is, pretty much by definition, incomplete. It’s really really hard to justify a claim that evolution cannot account for X, if you’re basing that on an incomplete model of evolution.

    The “rugged landscape” experiment clearly did use an incomplete model of evolution: just one type of mutation (single-base substitutions), selection for infectivity (with a fixed fitness landscape), and drift. I think you can make a good case that many other mechanisms can be discarded as irrelevant (e.g. anything that applies only to diploid organisms), but not all of them. The authors think the omission of “other mechanisms, such as homologous recombination” might be significant. You’ve portrayed this as a dishonest attempt on their part to avoid admitting that they’ve falsified evolution, but I can’t see any basis at all for your characterization.

    Even if they hadn’t stated it in the paper, it’d still be true. Every analysis (of real phenomena) has limitations, and in reasoning from them it’s important to understand those limitations and how they affect the conclusions you’re drawing. It’s also a bad idea to take any single experiment or analysis as the final word on anything. Published papers sometimes turn out to be seriously wrong, often contain minor mistakes, and even when things go “perfectly”, there are still things like measurement and statistical errors to contend with.

    Has their experimental result been replicated? Have they or anyone else done any testing to see how well the theoretical model they use corresponds to reality? How much uncertainty is there in the fitting from their data to the parameters of the theoretical model, and how does that uncertainty propagate through the model to the conclusion?

    Actually, let me expand on that last point a bit. If this were a physics paper(*), I’d have expected to have seen some discussion of the goodness-of-fit measure they used to fit the theoretical model to their data, and rather than just a single best-fit set of parameters, an “allowed region” of combinations of model parameters that gave acceptable fits. Then, I’d expect them to track how that range of parameters propagates through the model into its outputs. Basically, I expect to see error bars (or more complicated error range descriptions) everywhere. But the only place I see them is on the measured fitness results.

    (* I’m not a physicist, but my father was, I studied it quite a bit in college before deciding to pursue computer science instead, and I still find physics interesting so I still follow the subject a fair bit. Net result: I’m fairly familiar with how physics is done and how physicists think. Therefore, it tends to be my reference for how science is done and how to think about other scientific subjects.)

    Lack of error bars (or some other indication of estimated uncertainty) doesn’t mean there’s no uncertainty; quite the opposite, it means you don’t have any real idea how much uncertainty there is!

    BTW, you can get some idea how closely the model fits their data from Figure 3, which shows the measured fitness levels and fitted curve. Some points are very close to the curve, others pretty far away. Can you tell how well it matches with reality in the region they didn’t measure? I’m pretty sure the answer has to be “no”.

  36. 36
    gpuccio says:

    Gordon Davisson:

    Welcome back!

    I see the discussion is turning into a melee (sigh).

    Well, that seems unavoidable… 🙂

    I’m going to basically ignore everything since gpuccio @ 13,

    That’s very wise!

    and follow my original plan of discussing whether whether RM + NS is the official model of evolutionary theory (and then segue back to talking about the “rugged fitness landscape” paper).

    Perfect.

    OK, now I am busy, and cannot answer in detail. I’ll do it later. But I want to state immediately that I find your post #35 very reasonable, and that I can agree with many of the things you say there. Of course, with different emphasis and perspective, in many cases.

    So, it’s a pleasure to discuss again, and with a good interlocutor, after the unavoidable plunge into the melee (which sometimes is some experience, too! 😉 )

  37. 37
    Dionisio says:

    gpuccio,

    I’m glad to follow closely your interesting discussion with GD, but will try to stay on the sideline, so that I don’t get accused of turning this thread into a mêlée. 🙂

    Maybe my off topic comment @28 added up to the alleged mêlée? Sorry if that’s the case.

    Now just wanted to share this link before running to the exit door:

    http://onlinelibrary.wiley.com.....13047/full

    That’s all. No more unsolicited commenting here, unless I forget my promise. 🙂

  38. 38
    gpuccio says:

    Dionisio:

    “Maybe my off topic comment @28 added up to the alleged mêlée? Sorry if that’s the case.”

    No, I think he was probably referring more to the kind interactions between Corey Delvine, ET and me.

    “No more unsolicited commenting here, unless I forget my promise.”

    Please, forget, please, forget!

    By the way, I must apologize for my unsolicited summoning you in my answer to Corey at #32 (the emphasized part toward the end, about understanding evolution).

    I know, I should not have abused of your identity in that blatant way, but it was too good a temptation, and I could not resist! 🙂

  39. 39
    Dionisio says:

    gpuccio,

    When I read @32 the emphasized quote of your interlocutor referring to those who don’t understand evolution and your response about not being surprised by what your interlocutor had said, I couldn’t stop laughing even after reading your request to not laugh so loud. Well, when I read your request I laughed even louder. Sorry for that. Your interlocutor left me without choice.
    Laughing was the only alternative. Alright, I’ll try and behave better next time.
    I’ve been told more than once that I don’t have good manners in public forums. 🙂
    Actually, a Canadian professor said publicly here that I don’t ask honest questions, whatever that meant. 🙂
    But I want this serious discussion between GD and you to continue smoothly, because I expect to learn from it.
    I’ll walk out of the room next time your interlocutors write something that makes me laugh. 🙂

  40. 40
    gpuccio says:

    Gordon Davisson:

    OK, the first point:

    1) Is there an “official model” od evolution?

    Trying to define an “official model” of evolution is a lot like trying to define one for genetics or chemistry or geology. All of these study complex, diverse, and messy categories of phenomena; they’re fields that are constantly in flux, with new things being learned and sometimes things we thought we knew being reassessed. Trying to nail down exactly what the state of knowledge of any of these fields at any given time would be an exercise in nailing jello to a tree. What you can say, at any given time, is that there’s general agreement about some core knowledge about the phenomena, surrounded (metaphorically) by a cloud of reasonably-well-supported hypotheses, which are surrounded in turn by many many open and/or controversial questions. But the question of how settled various questions are is itself subjective, so you can’t even objectively define how solid various parts of our understanding are.

    This isn’t anything special about evolution; it’s just how things are in any dynamic field of science.

    Well, I find all that perfectly reasonable. I agree.

    Now, despite the difficulty of defining exactly what the official model/settled core/whatever of evolution is, I can confidently say that RM + NS is not it. At the very least, it needs to include genetic drift as a core process. But RM + NS + GD would still be a drastic oversimplification.

    OK, but here I must start to clarify what I think, which is slightly different, although not dramatically different.

    First of all, I must apologize for havinf used the rather inappropriate term “RM”, instead of the more accurate form “RV” (random variation), which I usually use. Probably, I did that because we were discussing the rugged landscape paper, which indeed can be more easily ointerpreted in terms of nucleotide mutations.

    However, I am well aware that there are many kinds of random variation events.

    Now, I have tried to explain, in my post #23 in answer to Corey Delvine, why I consider all forms of random variation events essentially equivalent (with some exceptions, that I have mentioned there) in terms of attempts at testing different states in the search space by a random walk. You may perhaps try to read that post, even if I happily admit it’s in some way part of the melee, because I believe there are some interesting points in it. 🙂

    So, form now on, it’s RV + NS.

    But, essentially, you are no, like Corey, objecting to the simplification in the types of variation. You are rather objecting, if Iunderstand well, to the simplification in other mechanisms that contribute to evolutionary history. First of all genetic drift.

    In that, you are probably right. For example, there is no doubt at all that neutral mutations and genetic drift are an important part of what happens in evolutionary history.

    So, did I mean when I said:

    “Well, they [the authors of the paper] didn’t use “a simplified model of evolution”. They tested the official model: RM + NS. And it failed!”

    I maintain that statement, but I realize that I meant something different by the term “official model” than what you mean with the same words.

    Let’s see. I agree with you that having some detailed model of what happened in evolutionary history is an ongoing task, and that there are a lot of different approaches to that, and different models, including, I would say, different design models which, although ognored by the official academy, are part of the scientific thought just the same.

    But in my statement I was not referring to a general model of evolutionary history, but rather to some explanatory model of the generation of new comples functional information, which, as you may have understood now, is my real reference. IOWs, we are not discussion “evolution” here, but rather the generation of new complex functional information. At least, that’s my view of things.

    So, the simple point is that I am convinced that, when we consider carefully what has been suggested in the last 70 years, and all the knowledge that has been accumulated in biology, the simple fact remains, IMO, that there are still only two types of “explanatory models” which have some potentiality to explain complex functional information in biological objects. And they are:

    a) The standard neo-darwinist model (RV + NS)

    b) Design models

    What do I mean by “standard neo darwinist model”? Very simply, the modern synthesis, in all its formulations, past and present.

    Because, in all its “official” formulations, the fact remains that the generation of new complex functional information is always explained as a result of RV + NS, and nothing else.

    Let’s consider, for example, the most important “addition” to standard evolutionary thought: neutralism, and the concept of genetic drift, as you very correctly mention.

    I have nothing against neutralism and genetic drift. I am sure that neutralists are very right, and that genetic drift happens all the time.

    But the simple fact is that those aspects are irrelevant in regard to the generation of new, complex functional information. They have no relevant influence on the probabilistic problem which is the core of ID thought. IOWs, they can considered essentially as variations of RV.

    GD, for example, can fix any new variation. But any new variation has the same probability of being fixed. Nothing changes.

    Indeed, even convinced neutralists, like Larry Moran, when brutally asked something about how relevant functional complexity is generated, have to admit (rather reluctantly, IMO), that the explanation for that trivial aspect can only be RV + NS! 🙂

    So, my view is that GD is an important part of what happens in natural history, but it is no part at all of any explicatory model for the generation of new complex functional information, simply because ot has no possible role in explaining it.

    Now, I am aware that there are other “models”, in what I often call “neo-neo-darwinism”, or “post-neo-darwinism”, that have the ambition to “explain” complex functional information without recurring to RV + NS as the onnly relevant mechanism.

    I have witnessed many of them, through the years, especially recently.

    Just as examples, let’s quote Andreas Wagner’s “Arrival of the fittest:

    “Natural selection can preserve innovations, but it cannot create them. Nature’s many innovations—some uncannily perfect—call for natural principles that accelerate life’s ability to innovate, its innovability.”

    and so on…

    and James A. Shapiro’s “Third way to evolution”:

    “James A. Shapiro proposes an important new paradigm for understanding biological evolution, the core organizing principle of biology. Shapiro introduces crucial new molecular evidence that tests the conventional scientific view of evolution based on the neo-Darwinian synthesis, shows why this view is inadequate to today’s evidence, and presents a compelling alternative view of the evolutionary process that reflects the shift in life sciences towards a more information- and systems-based approach in Evolution: A View from the 21st Century.”

    and so on…

    Well, my “not so humble” opinion about those models is simple: they explain nothing.

    Maybe I am wrong. But I don’t think so. Time will say.

    In the meantime, the only real “antagonist” to design models, IM not so humble O, remains one and only one: RV + NS.

    Why?

    Because RV + NS does explain something: for example, the known cases of microevolution.

    And because RV + NS is supported by some facts: for example, the known cases of microevolution.

    OK, a little boring maybe, but at least it is true.

    What RV + NS cannot explain at all is the generation of new, complex functional information. Which is the only thing I am interested in. Thats’ why I have dedicated my discussion with you, and this OP with the following comments, essentially to that point.

    Then you say:

    Really, it’d need to consist of a catalog of all of the various evolutionary mechanisms we’ve discovered over the decades. Most of them can at-least-sort-of be categorised as types of mutation (e.g. unequal crossing over, transposon and retroviral insertion), selection (biased gene conversion, meiotic drive), or genetic drift (the founder effect, hitchhiking), but not always, and not always into just one category. The hypermutable state model, for example, involves an unusual type of mutation tightly coupled to an unusual type of selection, so it sort of fits into both of those categories. On the other hand, endosymbiosis and homologous recombination don’t really fit into any of the categories.

    Any model of evolution that doesn’t include all of the above mechanisms (and others I didn’t list / don’t even know about) is, pretty much by definition, incomplete. It’s really really hard to justify a claim that evolution cannot account for X, if you’re basing that on an incomplete model of evolution.

    OK, I have probably already answered that. You are right if you speak of “evolution” as “anything that happend in evolutionary hostory”.

    But, if we are speaking of potential explanation, our choice in our debate is drastically restricted to what has been suggested and has explanatory power.

    IOWs, I am comparing the design model to any posiible explanation that exists in present scientific thought, not to any possible imagined future explanation of which we have no idea at all.

    And I stick to my idea that the only the RV + NS algorithm adds something to the mere idea of random events, and therefore deserves to be analyzed in the debate about design models.

    But, of course, if you or any other discussant can show me why other ideas really have some explanatory in regard to the generation of new complex functional information in biology, I ma ready to listen. And to answer. Just give me something that can be understood and answered!

    For example, you mention the “hypermutable state model”. I checked that briefly, and I found this paper:

    “Regulating general mutation rates: examination of the hypermutable state model for Cairnsian adaptive mutation.”

    Abstract
    In the lac adaptive mutation system of Cairns, selected mutant colonies but not unselected mutant types appear to arise from a nongrowing population of Escherichia coli. The general mutagenesis suffered by the selected mutants has been interpreted as support for the idea that E. coli possesses an evolved (and therefore beneficial) mechanism that increases the mutation rate in response to stress (the hypermutable state model, HSM). This mechanism is proposed to allow faster genetic adaptation to stressful conditions and to explain why mutations appear directed to useful sites. Analysis of the HSM reveals that it requires implausibly intense mutagenesis (10(5) times the unselected rate) and even then cannot account for the behavior of the Cairns system. The assumptions of the HSM predict that selected revertants will carry an average of eight deleterious null mutations and thus seem unlikely to be successful in long-term evolution. The experimentally observed 35-fold increase in the level of general mutagenesis cannot account for even one Lac(+) revertant from a mutagenized subpopulation of 10(5) cells (the number proposed to enter the hypermutable state). We conclude that temporary general mutagenesis during stress is unlikely to provide a long-term selective advantage in this or any similar genetic system.

    If I understand well, whether this model is good or bad, it is based on the general idea that some biological beings (especially bacteria, I believe) have adaptive mechanisms that can help the search for some new functional information (always, however, of the microevolutionary type).

    But I agree with that! I am convinced, for example, that the whole plasmidic system in bacteria is strongly adaptive.

    And we have a special example of that kind of adaptation in humans, which works on somatic immune cells and not on germ cells: the antibody maturation process, which I have discussed in detail here:

    https://uncommondescent.com/intelligent-design/antibody-affinity-maturation-as-an-engineering-process-and-other-things/

    The problem is that adaptive systems require a lot of complex functional information: they are indeed examples of objects with a huge amount of specific complex functional information. The advantage they can get in a random search is traceable to their functional information, which makes the random serach less random. We are, here, exactly in the context of the principle of conservation of information, many times discussed by Dembski and Marks.

    It’s like designing a computer which makes some specific random searches easier and more profitable. Wonderful but that’s the result of good design. You get what you inputted first, in a different form.

    OK, so I am here, ready to discuss any model which does not rely exclusively on RV + NS and which, according to what you believe, has some potential explanatory power in regard to the generation of new complex functional information in biology.

    In next post, I will discuss the rugged landscape paper. Again! 🙂

  41. 41
    gpuccio says:

    Dionisio:

    “A good laugh is a mighty good thing” 🙂

  42. 42
    ET says:

    Gordon:

    my original plan of discussing whether whether RM + NS is the official model of evolutionary theory

    First you have to find this alleged evolutionary theory. Then you have to read it to see what it actually says.

    Good luck with that

  43. 43
    Dionisio says:

    The explanation @40 could be a separate OP by itself.

    Like the bigos*, which tastes better the longer it lasts, this discussion is getting better with every hour.

    Kudos to both GP and GD.

    Keep it going!

    PS. I’m chewing slowly what is being said here. I want to digest it well. 🙂

    (*) https://en.wikipedia.org/wiki/Bigos

  44. 44
    Mung says:

    Gordon:

    I’m really good at procrastinating!

    But to be really good, like I am, you have to be good at procrastinating your procrastination!

  45. 45
    tribune7 says:

    –“official model” of evolution is a lot like trying to define one for genetics or chemistry or geology.–

    One can take issue with whether evolution should be compared to genetics, chemistry or geology, but leaving that aside those fields have “official models” i.e. standards.

    What would be the standard for evolution?

  46. 46
    gpuccio says:

    Mung:

    “But to be really good, like I am, you have to be good at procrastinating your procrastination!”

    I am an exceptional case, I must say: I always procrastinate, except when I don’t procrastinate.

    (See how spending time with neo-darwinists can teach us a lot! 🙂 )

  47. 47
    Dionisio says:

    gpuccio @46:

    “See how spending time with neo-darwinists can teach us a lot!”

    Could that qualify as beneficial mutation? 🙂

  48. 48
    Dionisio says:

    gpuccio @40:

    Shouldn’t any model of evolution include the core formulation of the fundamental evo-devo problem?

    Dev(d) = Dev(a) + Delta(a,d)
    Where Dev(a) is the entire developmental process of an ancestor, Dev(d) is the entire developmental process of a descendant, Delta(a,d) is the union of all the spatiotemporal changes in Dev(a) required to get Dev(d).

    Please, note that “entire” means the whole enchilada. tutto, tutto, tutto.

    However, a major outstanding issue is that we don’t understand the Dev(x) of any biological system ‘x’ yet. Work in progress… stay tuned.

    But the more we understand about it, the more it looks like designed systems.

  49. 49
    gpuccio says:

    Dionisio:

    “Could that qualify as beneficial mutation? ”

    I would rather say: environment driven adaptation (probably epigenetic). 🙂

  50. 50
    gpuccio says:

    Dionisio at #48:

    The whole enchilada, yes. And a rich enchilada it seems to be! At any level.

    “But the more we understand about it, the more it looks like designed systems.”

    There is no doubt about it. And if you consider that, to me, it already looked like designed systems, say 10 years ago, and irrefutably so, you can imagine how great is my conviction now!

    Irrefutable would be a pale euphemism.

  51. 51
    Origenes says:

    GPuccio,

    Would you agree when I say that both ‘negative selection’ and ‘positive selection’ results in loss of biological information? Both negative and positive selection results in the removal of certain organisms from the population.

    Now, if I am correct, the question arises: how does information loss help evolution?

    Those removed organisms contained unique information that is lost forever. It may even be the case that a removed organism was one or two point mutations away from some astounding evolutionary novelty, but due to actions of ‘positive selection’ the world will never witness it. What a horrible thought! 🙂

    Which brings me to a more general claim: natural selection (positive or negative) leads to loss of information and is therefore detrimental to evolution.

    Do you agree?

  52. 52
    Dionisio says:

    Origenes @51,

    gpuccio is more qualified to answer your questions about a topic I don’t understand well, but there’s something very basic about any selection: it must have things to select from to begin with.

  53. 53
    gpuccio says:

    Origenes:

    “Would you agree when I say that both ‘negative selection’ and ‘positive selection’ results in loss of biological information? Both negative and positive selection results in the removal of certain organisms from the population.”

    That’s true, of course. But let’s see more in detail the various scenarios.

    a) Negative selection. I would say that this results in a defense of the existing information (certainly not in the generation of new information).

    Of course, the cost of it is the loss of any new information in the individuals that are eliminated. Therefore, it is definitely against evolution by random variation, because any random variation that is a first step toward something new, but reduces the existing function, will be selected against.

    Moreover, the individual being selected against could have been hosting other potentailly beneficial variations, and those too will be lost. This is probably a very minor effect, but it is real.

    So, for negative selection you are certainly right.

    b) For positive selection, the issue is more tricky. We have to consider that expanding the new beneficial trait (whatever it may be), is and when it happens, can be considered a small gain in functional information in the population, because without positive selection the “beneficial mutation” would remain confined to a very small part of the population, and probably be lost to genetic drift.

    But, on the other hand, the elimination of the old population with the old trait is certainly a loss of information, in at least two important senses:

    b1) First of all, it is possible, and often true, that the new “beneficial” mutation is beneficial only in a transitory way, because of some strong selective pressure from the environment, and it could imply some loss of function in regard to the original trait. That can be the case in some forms of antibiotic resistance, for example. Or in some of the variations observed in Lenski’s experiment. So, the originaol trait could be the more functional in a standard environment, and if it is completely erased by positive selection uder extreme pressure, that useful information will be definitively lost.

    b2) But the most impportant aspect of information loss by positive selection is Haldane’s dilemma. Is two or more beneficial trait have to expand at the same time in a population, from different parts of it, there will be competition: the expansion of one trait may mean the loss of the others.

    So, you are completely right about negative selection. You are partly right, IMO, about positive selection, because we have to ackowledge that, in some cases, for example in some microevolutionary contexts, the expansion and fixation of a new beneficail trait is a gain of functional information.

    The problem is that positive selection is really rare, especially at molecular level, while negative selection is all pervasive.

    And beneficial mutations are extremely rare, while negative or neutral mutations are all pervasive.

    That does not sound good for neo-darwinism, does it? 🙂

    But, in the end, the strongest argument against the power of positive NS remains, IMO, the one that I have tried to emphasize throughout this thread:

    a) Positive Natural Selection can only act on what exists, and the supposedly beneficial variations that arise by chance are always informationally simple.

    b) New complex functional information cannot be deconstructed into simpler steps, each of them more functional and naturally selectable. For NS to have a chance with complex information, such a ladder of simple gradual selectable variations should exist for each example of complex functional information, IOWs for each existing protein. As far as we know, it exists for none.

    And there is no reason at all for such a ladder to exist. It does not exist for language, it does not exist for software, it does not exist for proteins.

    It does not exist for any functional unit where the function depends critically on the final configuration of many bits and is not present unless that functional configuration has been completely implemented.

    This is, IMO, the true and final argument against the power of NS to build up complex functional information. It simply can’t.

  54. 54
    Dionisio says:

    gpuccio @46:
    “See how spending time with neo-darwinists can teach us a lot!”

    Dionisio @47:
    “Could that qualify as beneficial mutation?”

    gpuccio @49:
    “I would rather say: environment driven adaptation (probably epigenetic).”

    Yes, agree, that’s much more evidence-based accurate. 🙂

  55. 55
    Dionisio says:

    gpuccio @53:

    “…for example in some microevolutionary contexts, the expansion and fixation of a new beneficial trait is a gain of functional information.”

    Is that the famous case of the Galapagos finch skull shape story?

    Weren’t those traits cyclically beneficial?

    One trait was negatively selected and another positively selected under strong selective pressures, but then in some cases things went back to square one under the opposite conditions and so on?

  56. 56
    Dionisio says:

    #55 addendum:

    Did those beneficial traits result from variations to existing genetic/epigenetic states that were relatively easily changeable without having major effect on the kind of biological system?

    How did some of those changes turn cyclical? Was it because the trait that had been previously eliminated never got 100% wiped out from the scenario or was it because the same changes were recreated?

    Please, feel free to correct the questions if they don’t make sense.

    Thanks.

  57. 57
    Origenes says:

    GPuccio @53

    Origenes: natural selection (positive or negative) leads to loss of information and is therefore detrimental to evolution.

    GPuccio’s answer:

    … you are completely right about negative selection.

    Good to know that you agree! Moreover ….

    positive selection is really rare, especially at molecular level, while negative selection is all pervasive.

    So, not only am I completely right about negative selection, but negative selection is by far the prevalent form of natural selection. Therefore I am 99% right. Do you hear that my evolutionist friends?

    However,

    GPuccio: You are partly right, IMO, about positive selection …
    We have to consider that expanding the new beneficial trait (whatever it may be), is and when it happens, can be considered a small gain in functional information in the population, because without positive selection the “beneficial mutation” would remain confined to a very small part of the population, and probably be lost to genetic drift.

    So, positive selection, on its own, not considering other factors such as drift, reduces information. This, on its own, is bad for evolution because it reduces the search capability of evolution. Okay. But that’s not the whole story … there is another competitive ‘information reducer’ in play, namely genetic drift, and to let this force win is an even worse option from an evolutionary standpoint. So, positive selection is bad for evolution, but because it is not quite as bad as genetic drift it is, in a sense, actually … “good”.
    The lesser of two evils, between a rock and a hard place or between Scylla and Charybdis.

    – – –
    I will respond to the remainder of your post later.

  58. 58
    gpuccio says:

    Dionisio:

    “Is that the famous case of the Galapagos finch skull shape story?”

    No, I was thinking more of the classic cases of microevolution, like simple antibiotic resistance, or the relative expansion of drepanocytosis and thalassemia alleles in malaria zones.

    Maybe the emergence of nylonase from beta-lactames could figure too, even if I think that was a microevolutionary event in the context of the adaptive plasmidic system in bacteria.

    Regarding Galapagos finches, I never discuss phenotypic shifts whose molecular cause is not known. And I am not sure that the molecular cause of those shifts in finches is known. Frankly, I have never been interested in that topic, which often seems to be mere propaganda. First, we must know as well as possible what happens at the molecular level, then we can discuss phenotypes.

  59. 59
    Dionisio says:

    gpuccio,

    I see your point, which makes sense.

    Thank you.

  60. 60
    Origenes says:

    GPuccio @53

    I have attempted to formulate your argument against NS in premises and a conclusion. Let me know what you think.

    (1) For NS to be the cause of complex specified functional information (CSFI), a ladder of simple gradual selectable (functional) variations leading up to CSFI should exist.
    (2) We find CSFI in language, software code and DNA code.
    (3) There exists no ladder of simple gradual selectable (functional) variations for CSFI in language and software code.
    (4) We have not found a ladder of simple gradual selectable (functional) variations for CSFI in DNA code.

    Therefore (from 3 and 4)

    (c) We have no reason to believe that NS is the cause for CSFI.

  61. 61
    gpuccio says:

    Origenes:

    Yes, you have understood my argument perfectly. I am happy of that! 🙂

    Only one thing, just to be fastidious: I would change “selectable” with “naturally selectable”. In all statements.

    Just to be clear:

    a) “Functional” is not correct. Anything can be functional, because for practically anything we can define some function. And one of the fundamental aspects of my approach to funcrional complexity is that we are free to define any possible function for an object, and to measure the complexity linked to thet function.

    The details of my definition of functional information and functional complexity can be found in this OP:

    https://uncommondescent.com/intelligent-design/functional-information-defined/

    b) “Selectable” is not correct. Artificial (intelligent) selection can select any possible function.

    c) “Naturally selectable” is the correct, necessary term. Naturally selectable functions are only those that can give some detectable reproductive advantage, and therefore can be expanded in a population because of their intrinsic “value” in that context.

    There is a huge difference between natural selection and artificial (intelligent) selection.

    NS has very limited power, as discussed in this whole thread. Definitely, it cannot generate new complex functional information, of any kind.

    AS is much more powerful, because it benefits of the information inputted in the selection itself.

    The added information takes three different forms in AS:

    1) A specific function is defined in advance, and the search is modeled exactly to generate that function

    2) The function can be measure at any desired threshold, the lower the better, according to the information implemented in the measuring system

    3) The measured function is rewarded not because of any intrinsic value it has in the system, but only because it is “recognized” by the system as the pre-defined function to be found. IOWs, the rewarding takes place through a symbolic recognition. The rewarding itself, therefore, is not intrinsic in the system, but requires addition specific information implemented in advance

    I have discussed in detail the differences between NS and AS in this OP:

    https://uncommondescent.com/intelligent-design/natural-selection-vs-artificial-selection/

    That said, we can understand why AS is much more powerful than NS. AS, if well implemented, can find its function thorugh RV with some ease. Even apparently complex functional information can be found in that way, because the complex functional information is the result of the complex functional information already implemented in the system.

    For example, the famous (and infamous) paper by Szostak about finding an ATP binding protein in a pool of random sequences is, obviously, a paper about successful AS discuised as a paper about NS, as I have argued many times, almost always arising angry reactions from someone. 🙂

    Another, extreme example of successful AS is Dawkin’s famous “Methinks it is like a weasel” program. Which clearly demonstrates that, if we already know a phrase, we can extract it by gradual RV, using the phrase as an oracle for our AS. Brilliant, indeed! 🙂

    Even children know that they can find their treasure at a hot and cold game, if they play well.

    That’s why it is important to stick to the full concept of “naturally selectable” when we argue about neo-darwinism and its supposed powers.

    Darwinists have always tried to disguise artificial selection as natural selection, and they will go on doing so, because they have no arguments in favor of NS as an explanation for complex functional information, and they desperately need to invent something!

  62. 62
    gpuccio says:

    Gordon Davisson:

    The second point:

    2) The rugged landscape paper, again.

    You say:

    The “rugged landscape” experiment clearly did use an incomplete model of evolution: just one type of mutation (single-base substitutions), selection for infectivity (with a fixed fitness landscape), and drift.

    OK, maybe we can come to some understanding here.

    After the arguments in my post #40, I think I have made clear that we should not be interested in a model of evolution, but rather in a model of some specific explanation of how evolution generates complex functional information. Which is exactly what the authors have tried to effect.

    A lab experiment is always a simulation. It is not direct observation of what happens in the wild. Therefore, it can never be complete.

    But a simulation can try to imitate how a model works, as well as possible.

    Now, you say that this experiment included the following elements:

    a) just one type of mutation (single-base substitutions),

    b) selection for infectivity

    c) a fixed fitness landscape

    d) genetic drift

    That is correct. But we should understand better what was implemented by the authors, and what was simply observed in the system.

    a) Mutations were implemented by the authors. The method they used is called “error-prone PCR”. Here is a brief description of the method:

    Error-prone PCR (EP-PCR) is the method of choice for introducing random mutations into a defined segment of DNA that is too long to be chemically synthesized as a degenerate sequence. Using EP-PCR, the 5′ and 3′ boundaries of the mutated region may be defined by the choice of PCR primers. Accordingly, it is possible to mutagenize an entire gene or merely a segment of a gene. The average number of mutations per DNA fragment can be controlled as a function of the number of EP-PCR doublings performed. The EP-PCR technique described here is for a 400-bp sequence, and an Alternate Protocol is for a library. EP-PCR takes advantage of the inherently low fidelity of Taq DNA polymerase, which may be further decreased by the addition of Mn2+, increasing the Mg2+ concentration, and using unequal dNTP concentrations.

    (Abstract from “Random mutagenesis by PCR.”):

    http://onlinelibrary.wiley.com.....6A7.f02t02

    So, this is a good point in the simulation, because the authors use an artificial method which can be considered an acceptable simulation of single mutations that occur in nature as the result of errors in DNA duplication. That is good methodology.

    b) Selection for infectivity. That, again, was done by the authors. So, in a sense, this is Artificial selection, not Natural Selection.

    But here the methodology is correct. The authors chose to select for infectivity, which in phages is essetnially the same thing as reproductive power. So, this type of srtificial selection for the same property that would allow natural selection in the wild is appropriate, and it can be considered an acceptable simulation for NS.

    (this is completely different from what happened in the famous Szostak experiment, which I will consider later).

    c) Fixed fitness landscape and d) Genetic drift

    These are not elements simulated by the authors, but rather the natural response of the experimental set to the simulation.

    I don’t think that the fixed fitness landscape is really different from many natural settings: all epxeriments about antibiotic resistance, for exmaple, can be considered a fixed fitness landscape in very much the same way. And antibiotic resistance is a perfectly natural example of neo-darwinian mechanism, often invoked as one of the (very few) examples of the powers of NS.

    So, if such a fixed landscape is good for malaria resistance reasonings, and even for Larry Moran, I would say it can be good for us too.

    Genetic drift occurred naturally in the system, and was not simulated at all. therefore, it is, I would say, a natural element in the system.

    So, we go back to the first point: they only used single mutations, or at least the kind of mutations that can arise in error-prone PCR.

    Why that? I don’t think there was any intention to use a “simplified model”. The simple truth is that sucvh a kind of artificial random mutations is probably the best we can do to simulate RV in the lab.

    Is this a severe limitation of the experiment?

    All depends on what you believe about the relevance of different types of variation as explanations of the generation of new complex functional information.

    As I have tried to argue, in regard to the generation of new complex functional information, all kinds of ravdom variation events are essentially the same: tghey just test one single state in the search space.

    Here is what I wrote to you in my #40:

    “However, I am well aware that there are many kinds of random variation events.

    Now, I have tried to explain, in my post #23 in answer to Corey Delvine, why I consider all forms of random variation events essentially equivalent (with some exceptions, that I have mentioned there) in terms of attempts at testing different states in the search space by a random walk. You may perhaps try to read that post, even if I happily admit it’s in some way part of the melee, because I believe there are some interesting points in it.”

    However, if you really believe that the specific type of random variation event can change significantly the results, you should try to find somne support for that idea in the experimental literature. I would be happy to consider any examples of that.

    In the meantime, I would like to briefly compare our paper to another famous paper, the one by Szostak about a functional ATP binding protein from a pool of random sequences.

    I have said many times why that paper is methodologically inappropriate as a simulation of NS.

    The main reason, of course, is that it uses AS as a simulation of NS. But it is not AS for an equivalent of what would ne naturally selected, like in the rugged landscape paper (where infectivity is selected).

    It is, instead, AS for ATP binding.

    It is important to observe that the orinal sequence in the original random pool presented only very weak ATP binding, and certainly was not naturally selectable.

    But, even more strikingly, the final protein with strong ATP binding still was not naturally selectable at all.

    So, why am I citing yjos bad paper in relation to the good paper of rugged landscape?

    Because of course it not only contains all the mistakes I have already described which are not present in the rugged landscape paper, but it also shares with the rugged landscape paper the “limits” that you lament:

    a) Mutations are implemented by error-prone PCR, exactly like in the rugged landscape paper. No other source of variation is tested.

    b) The fitness landscape is absolutely fixed, indeed completely innatural (the proteins were selected by measuring their binding to ATP columns).

    Moreover, for the same reason, there was no genetic drift at all in the system.

    So, a paper which not only is inappropriate in its chosen methodology, using a completely wrong simulation of NS, but also shares the obvious limits which are probably shared by any current lab experiment about NS.

    So, what was the result of that in the academic response? Was the paper criticized for its inappropriate methodology? Were the limits in the basic experimental settings clearly outlined by scientists?

    Not at all!

    The paper was, and is, considered as a champion of neo-darwinism, the clear evedence of what NS can do, of how “functional” sequences can be easily found in a pool of random sequences.

    Its supposed (and completely wrong) conclusion, that you can find functional sequences in about 1:10^10 random sequences, has even been used to counter the correct conclusions by Axe that we need about 10^70 sequences to find a folding protein. In a famous darwinist site, the scenario was simply described as experimental results giving the frequency of functional sequences in a range as great as 10^10 – 10^70!

    OK, I will not say more, for the moment, about the “limits” of experimental results, and about how they can be intepreted as convenient, if the official dogma needs to be defended.

    (Beware, this discourse is not intended for you, of course. I have full certainty of your good faith and honesty. But my point is that you seem to share the general bias, of course unintentionally)

    More in next post.

  63. 63
    Origenes says:

    GPuccio @61, @53

    Thank you for your clear well-argued explanation.

    So, just for the record:

    (1) For NS to be the cause for complex specified functional information (CSFI), a ‘ladder of simple gradual naturally selectable variations’ leading up to CSFI should exist.
    (2) We find CSFI in language, software code and DNA code.
    (3) There exists no ‘ladder of simple gradual naturally selectable variations’ for CSFI in language and software code.
    (4) We have not found a ‘ladder of simple gradual naturally selectable variations’ for CSFI in DNA code.

    Therefore (from 3 and 4)

    (c) We have no reason to believe that NS is the cause for CSFI.

  64. 64
    gpuccio says:

    Origenes:

    Yes! 🙂

  65. 65
    gpuccio says:

    Gordon Davisson:

    You say:

    The authors think the omission of “other mechanisms, such as homologous recombination” might be significant. You’ve portrayed this as a dishonest attempt on their part to avoid admitting that they’ve falsified evolution, but I can’t see any basis at all for your characterization.

    Well, my point is not really to characterize the authors as “dishonest”: I have already said that I appreciate their work.

    But the way they express their conclusion is certainly incorrect.

    Look, again, I will try to say how that idea could have been expressed correctly:

    “Our data and our model have shown a significant and unexpected limit in the power of random variation and natural selection to find the wildtype from of the protein, because a library of 10^70, which according to our model would be requested to attain that result, is completely out of the range of resources of our planet and of the biological world. However, we have tested only one form of random variation, single nucleotide subdtitutions. It is possible that other kinds of variation, like recombination or others, could change the results. The role of recombination has been indeed suggested. Unfortunately, we can draw no conclusions about that, because we did not include recombination, or other types of variation, in our simulation. Further research is certainly needed to assess that important point.”

    Well, that would have been correct.

    Instead, what did they write?

    By extrapolation, we estimated that adaptive walking requires a library size of 10^70 with 35 substitutions to reach comparable fitness. Such a huge search is impractical and implies that evolution of the wild-type phage must have involved not only random substitutions but also other mechanisms, such as homologous recombination.

    (emphasis mine)

    And what is the result?

    The result is that when I cite correctly the conclusions derived from their data, that single substitutions and natural selection cannot find the wildtype, darwinists immediately say:

    “You are wrong. The authors say that recombination did it!”

    Including you.

    That’s what incorrect conclusions can achieve!

    More in next post.

  66. 66
    gpuccio says:

    Gordon Davisson:

    You say:

    Even if they hadn’t stated it in the paper, it’d still be true.

    What would still be true? That “the evolution of the wild-type phage must have involved recombination”? Who says that? What facts support that statement? Is this some dogma, or what?

    You say:

    Every analysis (of real phenomena) has limitations, and in reasoning from them it’s important to understand those limitations and how they affect the conclusions you’re drawing. It’s also a bad idea to take any single experiment or analysis as the final word on anything. Published papers sometimes turn out to be seriously wrong, often contain minor mistakes, and even when things go “perfectly”, there are still things like measurement and statistical errors to contend with.

    True. I agree. But that is true of any scientific model, indeed of any scientific knowledge.

    However, when I find a good paper which presents good data and reasonable intepretations, I give it serious consideration.

    Has their experimental result been replicated? Have they or anyone else done any testing to see how well the theoretical model they use corresponds to reality

    I don’t know. But I think that darwinists should be very eager to falsify those results, if they really understand what they mean. So, why don’t they do it?

    The paper is an important one. the results are very interesting. The paper was published in 2006.

    More in next post.

  67. 67
    gpuccio says:

    Gordon Davisson:

    Finally, you present some detailed criticism about the model itself. However, you don’t really have any definite argument against it, you just doubt that it can be considered reliable.

    Well, as I have alredy said, I cannot relly go into details about the model, because that requires, IMO, a level of mathematical competence that I don’t have. We should be able to read and understand well, at least, the:

    “Brief Introduction to the n-k Fitness Landscape”

    in the materials and methods section.

    I will not even try that. 🙂

    Regarding your objection to the fit shown in Fig. 3, I understand that you would like some p-value for the goodness of fit of the curve. So would I.

    But I don’t think that the points shown in the figure are the whole stuff (they are only mean fitnesses). The curve was probably computed on the raw data. And it does not seem a bad fit, looking at the figure, at least IMO.

    However, my little understanding of the model itself, and the lack of more details about the raw data, prevent me from being able to judge the statistical aspect (about which I would probably be more competent).

    So, I can only say again: I accept the model as it is, with the caution that we always must have about any scientific model, and I have no specific reason to think that it is wrong or grossly imprecise.

    However, your insistence in doubting those results is IMO some evidence that you have really understood their meaning.

    And that is really good! 🙂

  68. 68
    Dionisio says:

    gpuccio @61:

    “Even children know that they can find their treasure at a hot and cold game, if they play well.”

    Nice example of goal-based guided search.

  69. 69
    Dionisio says:

    gpuccio @66:

    Who says that?

    Well, the authors did. Isn’t that enough credibility? At least they understand evolution and we don’t. 🙂

    What facts support that statement?

    Huh? Facts? What’s that? 🙂

    Is this some dogma, or what?

    Maybe it is, why not? 🙂

  70. 70
    Corey Delvine says:

    Gpuccio, I seriosuly doubt that you have the support you claim.

    “a) I claim that “negative selection works against evolution.” Wikipedia seems to agree. Biologists seem to agree.”

    If you can find me a single source that says anything along the lines of “negative selection works against evolution”, I would be extremely surprised.

    Also, you want a discussion about negative selection acting in favor of, or neutral toward, evolution?
    Ok, maybe you just haven’t actually read a word I’ve said (I guess that wouldn’t really surpirse me), but here goes nothing:

    Negative selection removes deleterious mutations.
    This means that at the very least, it will maintain the current fitness state of a population, and therefore be neutral.
    So, at the very least, negative selection is neutral.
    That is definitely not “working against evolution”.
    Now, because this process helps to maintain the current fitness state, I am also arguing that negative selection ensures that other processes will shift the population toward a “more fit” state.
    By culling deleterious mutations, negative selection constantly holds a population in a state that is primed for further adaptation….evolution.

    You hem and haw about forests and trees, but you are apparently still missing the point.
    Negative and positive selection have different effects at the amino acid sequence level, but if you take a step back you’ll see that they are both processes occuring during evolution.
    In the end both processes contribute to the improvement of fitness for a species.
    You are obsessed with staring at amino acid sequences, but are missing the bigger picture.

    “Even one bit of information can crash a system”
    Sure, and with computers this is common.
    However, in biology there is almost always wiggle room.
    As I said, those examples were not about function.
    They were pointing out huge changes that can occur in an organism with little to no change in nucleotide sequence.
    And once again, your attempt to just sweep these under the rug do not look good on your part.

    Maybe you don’t need a few extra ribs, but think of the first organisms with only a partial rib cage.
    I’m sure they’d like a few more ribs to protect their vital organs.

    Or do you think every organism has and always had the same number of ribs as you?

    “show me how a simple change in the genetic code can generate a new complex proteins”
    Your question demonstrates you bias.
    Your use of the word new here I assume means you are asking for a sequence that is unlike other before it.
    However, you (I hope) and I both know that evolution does not claim to produce “new complex proteins” with a simple change.
    As I said, the original evolution of a protein and and the evolution of one protein from another are two completely different processes.

    Oh, you think those two are “maybe” different questions?
    “Maybe”?
    They are entirely different.
    The fact that you think they are only “maybe” different gives me a lot of perspective on you.

    And ET, ATP synthase acts as a catalyst and “speeds up” a chemical reaction just like any other enzyme.
    It does this through a transfer of energy, also just like any other enzyme.
    I’d call it quits if I were you.

  71. 71
    cmow says:

    I should probably stay out of this — gpuccio and ET can defend themselves, and this thread is really better as a discussion between gpuccio and Gordon…

    but… I’d say on the face of it, Corey Delvine, that you are making some bizarre statements.

    Like:

    Negative selection removes deleterious mutations.
    This means that at the very least, it will maintain the current fitness state of a population, and therefore be neutral.
    So, at the very least, negative selection is neutral.
    That is definitely not “working against evolution”.

    What? To ‘maintain’ is to cause a state of affairs to continue. To ‘maintain [a] current fitness state’ means continue the current state or ‘don’t change’, which means ‘don’t evolve’. In simple English, this quite literally is working against evolution.

    I think you’ve granted gpuccio’s points, but strangely your sensibilities seem to be telling you that you are winning a debate that you have stepped into ‘out of context’, apparently because gpuccio is not seeing the big picture. Odd.

  72. 72
    gpuccio says:

    cmow:

    Very good point, thank you! 🙂

    I will answer Corey Delvine’s comments in detail later (I have not the time just now), but I am happy that you have already said the first thing that came to my mind wnem reading his post.

    Thank you for the help. It is really appreciated. And there is absolutely no reason that you “stay out of this”, if you like to give some contribution! 🙂

  73. 73
    EugeneS says:

    GP

    Really interesting stuff. I have been busy with my commitments but I did notice that you addressed my question. Thank you very much. I will give this OP and the thread underneath a good read as soon as I can. Many Thanks!

  74. 74
    Origenes says:

    Corey Delvin @70

    If you can find me a single source that says anything along the lines of “negative selection works against evolution”, I would be extremely surprised.

    Are you familiar with neutral theory? It is claimed that this theory has become central to the study of evolution at the molecular level.
    This theory is based on the notion that selection works against evolution.

    There are many quotes to be found, see for instance here:

    It’s also true that negative natural selection acts as a break on evolution by preventing detrimental changes and “weeding out non-viable life forms.”

    [professor Larry Moran]

  75. 75
    Origenes says:

    // follow-up 74//

    Corey Delvin, I don’t want you to think that I could only find one quote by prominent evolutionists, so here are some more:

    many genomic features could not have emerged without a near-complete disengagement of the power of natural selection.

    [Michael Lynch, opening, The Origins of Genome Architecture]

    and

    … a relative lack of natural selection may be the prerequisite for major evolutionary advance

    [Mae Wan Ho, ‘Beyond Neo-Darwinism’]

    and

    The internal contradiction in its [natural selections’] major theoretical cornerstone — Fisher’s fundamental theorem … traits having been subjected to heavy selection pressures, because of their importance in the lives of the organisms, should be less variable than less important traits… . traits that have been most important in the lives of organisms up to this moment will be least likely to be able to evolve further!

    [Stanley Salthe, Critique of Natural Selection]

  76. 76
    gpuccio says:

    Origenes:

    Thank you! 🙂

    You have saved me a lot of work.

    Happy to be backed up by Larry Moran (and other important people), for once! 🙂

  77. 77
    gpuccio says:

    Corey Delvine:

    First of all, thank you for commenting. It seems that this time you try to make some arguments. That’s good. So I will happily answer you.

    If you can find me a single source that says anything along the lines of “negative selection works against evolution”, I would be extremely surprised.

    Well, it seems that Origenes has saved me the work.

    Also, you want a discussion about negative selection acting in favor of, or neutral toward, evolution?

    Yes. My appeal to arguments seems to have worked, then!

    Ok, maybe you just haven’t actually read a word I’ve said (I guess that wouldn’t really surpirse me), but here goes nothing:

    You are really wrong and unfair here. I have read alla that you have said with hreat attention, and answered point by point, in detail.

    Negative selection removes deleterious mutations.

    Correct! I appreciate that you seem to have accepted that negative selection and positive selection are two different processes, even if we agree that they share the eliminative character. That’s good.

    This means that at the very least, it will maintain the current fitness state of a population, and therefore be neutral.

    No, that’s wrong. First of all, if it maintains the “current fitness of the population” (which it does not always do, see diseases, or quasi neutral deleterious mutations), it certainly does not do “at the very least” that. It does “at most” that. Let’s try to use our words correctly.

    Moreover, “neutral” is a term that is used for mutations in relation to their effect on fiteness. Let’s not generate confusion. Negative selection is a process. Regarding evolution, the question here is if it acts in favor or against evolution, or if it is not involved in it. Are you saying that negative selection is not involved in evolution? I don’t agree, and I will explain why.

    So, at the very least, negative selection is neutral.
    That is definitely not “working against evolution”.

    Where is your logic? You are completely and blatantly ignoring the reason why negative selection acts against evolution. You are reaching conclusions without even starting a debate about the issue!

    Now, because this process helps to maintain the current fitness state, I am also arguing that negative selection ensures that other processes will shift the population toward a “more fit” state.
    By culling deleterious mutations, negative selection constantly holds a population in a state that is primed for further adaptation….evolution.

    As cmow has already brilliantly explained, you are making a serious error here.

    Nobody denies that negative selection (imperfectly) helps preserve the already existing functional information. I have said that many times, and very explicitly.

    If you want to argue that such a preservation is the premise for further evolution, I agree. But that does not mean that negative selection has any active role in promoting evolution, IOWs in the generation of new functional information. If it were for negative selection, information would at most remain the same as it already is.

    So, is it true, as you say, that it can be considered “neutral”, IOWs having no role in evolution?

    Absolutely not. As already said, the simple argument for the role against evolution played by negative selection is that it acts by opposing change, if change is deleterious to the existing function.

    That implies acting against evolution, because evolution need change, and in many cases that change will ne deleterious for the existing function.

    That’s why Wikipedia says:

    “The Ka/Ks ratio is used to infer the direction and magnitude of natural selection acting on protein coding genes. A ratio greater than 1 implies positive or Darwinian selection (driving change); less than 1 implies purifying or stabilizing selection (acting against change); and a ratio of exactly 1 indicates neutral (i.e. no) selection.”

    That’s why Larry Moran says:

    “It’s also true that negative natural selection acts as a break on evolution by preventing detrimental changes and “weeding out non-viable life forms.””

    Do you understand what “acting against change” means?

    OK, let’s go more in detail.

    Let’s imagine that some new trait cen be beneficial, and therefore contribute to evolution by positive selection, but only if two AA changes are present at the same time in an existing proteins. Let’s call the two substitutions A and B.

    Now, both A and B, individually, are slightly deleterious mutations (which is a very common case), but their mutual interactions makes them beneficial.

    This is a completely likely scenario, if you understand something of protein biochemistry.

    Now, if the probability of both A and B is, say, 10^-6, in some definite evolutionary time, then the probability of having both mutations in the same individaul in the same evolutionary time will be 10^-12: rare, but not impossible. It can happen.

    In the absence of any intervention by negative selection

    But if the deleterious effect of each individual mutation is strong enough to trigger negative selection, then the probability of having both mutations in a certain evolutionary time in the same individual becomes essentially zero, because whatever the first mutation is that takes place (A or B), it will be eliminated by negative selection.

    Is that clear?

    That’s why negative selection acts against evolution. I don’t believe that it’s a very strong effect, but it is completely true. Whatever you say.

    More in next post

  78. 78
    Origenes says:

    // follow-up #74 #75 //

    Professor PZ Meyers deserves a special mention:

    I think if selection were always the rule, then we’d never have evolved beyond prokaryotes — all that fancy stuff eukaryotes added just gets in the way of the one true business of evolution, reproduction…

    The bottom line is that you cannot easily explain most increases in complexity with adaptationist rationales. You have to consider chance as far more important, and far more likely to produced elaborations.…

    Even in something as specific as the physiological function of a biochemical pathway, adaptation isn’t the complete answer, and evolution relies on neutral or nearly neutral precursor events to produce greater functional complexity.

    // GPuccio @76

    At your service 🙂

  79. 79
    gpuccio says:

    Corey Delvine:

    You hem and haw about forests and trees, but you are apparently still missing the point.
    Negative and positive selection have different effects at the amino acid sequence level, but if you take a step back you’ll see that they are both processes occuring during evolution.
    In the end both processes contribute to the improvement of fitness for a species.
    You are obsessed with staring at amino acid sequences, but are missing the bigger picture.

    Again, this sounds only like a political slogan and nothing else. Certainly not like an argument.

    What do you mean when I say that I am “obsessed with staring at aminoacid sequences”?

    As I have already explained, I look at aminoacid sequences because:

    1) RV at the protein coding gene level changes nucleotide sequences that define AA sequences. That’s where the variation happens, in sequences of information.

    2) NS, both negative and positive, acts in the end on the genetic information present in sequences, by eliminating some sequences and preserving or expanding others.

    3) The whole neo-darwinian theory, therefore, is about information in genetic sequences.

    That’s why I look at sequences.

    What else should I look at?

    You say that I am “missing the bigger picture”? Do you know why that happens?

    Because your “bigger picture” is simply a brief collection of vague statements, which mean nothing and explain even less!

    Big picture! Please, explain how that picture works!

    You know the old saying: the devil is in the details.

    You are certainly free from the devil, I would say: no details at all in what you say. Just political slogans and propaganda.

    More in next post.

  80. 80
    gpuccio says:

    Origenes:

    Precious quote! Thanks agains. 🙂

    Two comments:

    1) I absolutely agree with PZ (a rare event indeed) on one point:

    “if selection were always the rule, then we’d never have evolved beyond prokaryotes ”

    Only I would say:

    if fitness were always the rule, then we’d never have evolved beyond prokaryotes.

    It is absolutely true that prokaryotes are by far the most successful kind of living beings, in regard to fitness and reproduction! As PZ says, no need for eukaryotes, least of all for metazoa. 🙂

    And I must say that flies, cockroaches and mice are great champions too.

    2) That said, I am sure that if anyone showed to PZ, or to any other “non selectionist”, some hard evidence of extremely high functional information for which there is no possible explanation (let’s say our classic beta chain of ATP synthase, just to make an example among thousands), in the end he would recur to fitness and NS again, since all explanations based on random variation alone would miserably fail.

    Please note the words I have emphasized here:

    “I think if selection were always the rule, then we’d never have evolved beyond prokaryotes — all that fancy stuff eukaryotes added just gets in the way of the one true business of evolution, reproduction…

    The bottom line is that you cannot easily explain most increases in complexity with adaptationist rationales. You have to consider chance as far more important, and far more likely to produced elaborations.…

    Even in something as specific as the physiological function of a biochemical pathway, adaptation isn’t the complete answer, and evolution relies on neutral or nearly neutral precursor events to produce greater functional complexity.”

    These fans of chance should really do some math about probability, a task they seem to abhor, instead of simply relying on the goold old NS in case of difficulties! 🙂

  81. 81
    gpuccio says:

    Corey Delvine:

    “Even one bit of information can crash a system”
    Sure, and with computers this is common.
    However, in biology there is almost always wiggle room.

    It’s rather common in biology too. Look at single gene disorders, for example.

    In many cases, a single mutation can completely erase the function of a protein. It depends on what is the aminoacid that changes, IOWs the functional information in that aminoacid site.

    Ah, but now you will accuse me again of being obsessed by sequences!

    As I said, those examples were not about function.
    They were pointing out huge changes that can occur in an organism with little to no change in nucleotide sequence.
    And once again, your attempt to just sweep these under the rug do not look good on your part.

    Of course they were not about function. They were about loss of function.

    So, what is your point? That a lot of function can be lost by some small change in information?

    I supposed that was my point!

    Not with “no change”, anyway!

    And I am not sweeping anything under a rug. I am only saying that new complex functional information cannot be generated with small changes in information. That’s all.

    More in next post.

  82. 82
    Dionisio says:

    cmow and Origenes

    Your willingness to assist gpuccio here is very commendable.

    As we all can see, gpuccio can use all the help he can get, because the poor guy doesn’t understand evolution.

    The problem is that we neither.

    We should learn biology 101 in order to understand what the Neo-Darwinian folks are trying so hard to tell us.

    🙂

  83. 83
    Dionisio says:

    gpuccio @77:

    “Let’s try to use our words correctly.”

    That’s not nice to ask your politely dissenting interlocutors to do something they usually don’t do.

    Please, be nicer next time.

    🙂

  84. 84
    Dionisio says:

    gpuccio @81:

    “Ah, but now you will accuse me again of being obsessed by sequences!”

    Well, you seem to be obsessed with complex functional specified information too. 🙂

    Also you seem to be obsessed with sudden appearance of large amounts of functional information in protein families. 🙂

    You may want to consult with a specialist that could treat your obsession problem before it gets out of control. 🙂

    Or better yet, try to understand evolution and thus all your obsession issues will disappear magically. Once you reach that level of understanding, you’ll be a happy camper, to whom nothing matters. Finally you won’t be concerned about proven facts or evidences.

    🙂

  85. 85
    Dionisio says:

    gpuccio @76:

    “Happy to be backed up by Larry Moran (and other important people), for once!”

    That’s because you’ve asked only honest questions. 🙂

  86. 86
    Dionisio says:

    gpuccio @81:

    “And I am not sweeping anything under a rug.”

    But you’re definitely sweeping and mopping the floor with your politely dissenting interlocutors’ weak arguments.

    🙂

  87. 87
    gpuccio says:

    Corey Delvine:

    Last points:

    Maybe you don’t need a few extra ribs, but think of the first organisms with only a partial rib cage.
    I’m sure they’d like a few more ribs to protect their vital organs.

    Or do you think every organism has and always had the same number of ribs as you?

    The exact molecular mechanisms that determine the correct number of ribs, or the correct size or form of parts of the body, are still poosrly understood. Maybe some of them are relatively simple at the molecular level, but I doubt it.

    As I have said at #58, I don’t discuss phenotypic aspects whose molecular foundation and control is not clearly understood.

    But, again, I do discuss molecular examples of comple functional information where the functional complexity is understood and can be measured. Like ATP synthase beta chain, just to go back to one classic example.

    “show me how a simple change in the genetic code can generate a new complex proteins”
    Your question demonstrates you bias.

    No, my question was motivated by your statement:

    Are you aware that deletion of a single HOX gene has been shown to shift the rib cage up and down the spine of model organisms?
    How about the fact that a chromosomal inversion in fruit flies drives growth of legs where antennae would normally be?
    While I am mentioning these more for shock value rather than actual biological function, these massive changes at the organismal level are driven by small or even no change in DNA sequence.
    Are you sure that you understand the breadth of what “random variation” entails?”

    As I fully understand the breadth of what destructive random variation, even simple, can entail, I wanted to clarify that, instead, constructive functional random variation, when simple, entails almost nothing.

    That’s why I asked my question.

    If you agree that simple random variation cannot entail any relevant complex result from the point of view of functional information, then we seem to agree.

    I will continue not to be amazed at the power of destructive random variation: whoever has destroyed a house of cards with a very slight movement knows that concept all too well.

    On the other hand, whoever has built a house of cards with a single, slight movement, is certainly a remarkable individual! 🙂

    More in next post.

  88. 88
    gpuccio says:

    Corey Delvine:

    As I said, the original evolution of a protein and and the evolution of one protein from another are two completely different processes.

    Oh, you think those two are “maybe” different questions?
    “Maybe”?
    They are entirely different.
    The fact that you think they are only “maybe” different gives me a lot of perspective on you.

    Well, giving perspective to others is always a good result, I suppose.

    Why maybe?

    I would not like to confuse your perspective, but the simple reason is that it is not clear what you mean.

    You say:

    the original evolution of a protein and and the evolution of one protein from another

    ???

    What is “the original evolution of a protein”?

    I think you do not mean “the evolution of the first protein”, do you?

    Now, we have about 2000 protein superfamilies (in the SCOP database), which are completely unrelated at the levels of sequence, structure and function.

    How did they originate?

    The common idea is that new proteins originate from existing proteins, by RV + NS. Do you agree?

    But then what is the difference between “the original evolution of a protein” and “the evolution of one protein from another”? I don’t understand.

    Let’s consider, again, the beta chain of ATP synthase, a 553 AAs long protein, with 334 identities between E. coli and humans. Lots of functional information here!

    Now, what is your model for the origin of that sequence? How was that highly specific sequence found?

    Did it come from some other existing protein?

    Or from what?

    Maybe with “the evolution of one protein from another” you mean only the cases where a small variation at the level of the active site can differentiate proteins inside an already existing family?

    If that is the case, please read what I said in my post #5, to EugeneS. I paste here the relevant part for your convenience:

    I think that the only reasonable scenario where a gene duplication could perhaps generate a new functional gene by the neo-darwinian mechanism is the following:

    A gene is duplicated, and remains functional. While the original gene ensures that the old function is satisfied, the new gene undergoes small variations, 1 – 5 AAs, at the active site, and is transformed into a similar gene, with more or less different biochemical activity.

    This could be a mechanism that generates diversification in an existing protein family, for example.

    The point is: most of the sequence and structure of the old gene, here, will be conserved. That’s why it is a good thing that the duplicated gene remains functional, so that negative selection can preserve that bulk of sequence, structure and functionality.

    On the other hand, while the folding and the general structure of the protein remain the same (IOWs, we remain in the original island of the original protein family) the active site can undergo some small variation that changes its biochemical affinity for substrates, and so in the end provides a different range of activity and function.

    This variation at the active site is usually in the microevolutionary range (as I said, 1 – 5 AAs), so it could be potentially in the range of very good biological probabilistic resources.

    So, what do I believe about this scenario? I believe that those cases are borderline: they could be extreme cases of neo-darwinian microevolution, or very simple cases of designed macroevolution.

    Therefore, it is wise IMO not to focus a design inference on that type of processes: we have indeed a lot of scenarios where the informational jump is hundreds of times bigger, beyond any possible reach of the neo-darwinian theory.

    For example, all cases of appearance of a new protein superfamily, or in general of a huge quantity of new information in a protein. You can take a look at my OPs about the informational jump in vertebrate proteome to find a lot of such examples.

    Now, the most revealing among your statements is perhaps the following:

    Your use of the word new here I assume means you are asking for a sequence that is unlike other before it.
    However, you (I hope) and I both know that evolution does not claim to produce “new complex proteins” with a simple change.

    Well, I use the word “new” because of course we are interested in new information, not only in the repetition of what already exists.

    It is not necessary, however, for the functional information to be new, that the sequence be completely “unlike other before it”.

    Of course, we have the 2000 protein superfamilies, already cited, which are a very good example of groups of proteins which are completely unrelated at sequence level, and also at structure level and function level.

    But, if you read my OP here:

    https://uncommondescent.com/intelligent-design/the-amazing-level-of-engineering-in-the-transition-to-the-vertebrate-proteome-a-global-analysis/

    you will see that I have shown that about 1.7 million bits of functional information (in the form of human conserved information) appear in the proteome at the transition from pre-vertebrates to vertebrates. Most of that new information appears in proteins that alredy existed in pre-vertebrates, abd which alredy exhibited some sequence homology with the human form of the proteins. But the fact is, a lot of new human conserved information is added in the transition to vertebrates.

    If you look at my OP here:

    https://uncommondescent.com/intelligent-design/interesting-proteins-dna-binding-proteins-satb1-and-satb2/

    you will see instead the example of a protein, SATB1, which practically does not exist in pre-vertebrates, and which appears in vertebrates practically almost identical to how it is in humans today: 1203 bits of homology, 79% identity.

    So, how do you explain those facts, and thousands like them?

    I have not seen, yet, any model presented by you.

    Unless we can consider a “model” your rich and amazing statement:

    “However, you (I hope) and I both know that evolution does not claim to produce “new complex proteins” with a simple change.”

    This from the same person who wrote (in post#26):

    “We did not walk our way to ATP synthase one amino acid at a time.
    No biologist would ever claim that’s how it happened.”

    So, let me understand:

    You say:

    “We did not walk our way to ATP synthase one amino acid at a time.”

    So, it was not a gradual walk. OK.

    But you also say:

    "However, you (I hope) and I both know that evolution does not claim to produce “new complex proteins” with a simple change.”

    So I am a little confused. How did it happen, according to your “model”?

    Was it a “complex” change, but not a gradual one?

    Was it magic?

    Did it appear out of thin air?

    Just to understand your position.

    OK, that’s all, for the moment.

  89. 89
    ET says:

    Corey:

    And ET, ATP synthase acts as a catalyst and “speeds up” a chemical reaction just like any other enzyme.
    It does this through a transfer of energy, also just like any other enzyme.

    No other enzyme works by force of pressure. As Larry Moran said the ADP and Pi are squeezed together. If you think I am wrong then go ahead, show me.

    See there isn’t any natural attraction to cause a reaction between the two. The only reason it happens is because it is squeezed on- pressure does it.

    But that is besides the point which is there isn’t a non-telic explanation for the existence of ATP synthase. So you lose regardless of any semantic mistake you think I have made.

    Without ATP any proto-organism would have to be tethered to an energy source until a mechanism to make ATP could just happen. But that begs the question- how did blind and mindless processes figure out that ATP would be the best currency for life?

    The “theory” of “it just happened” isn’t much help, is it?

  90. 90
    Dionisio says:

    gpuccio @87:

    The exact molecular mechanisms that determine the correct number of ribs, or the correct size or form of parts of the body, are still poorly understood. Maybe some of them are relatively simple at the molecular level, but I doubt it.

    As I have said at #58, I don’t discuss phenotypic aspects whose molecular foundation and control is not clearly understood.

    That makes much sense.

  91. 91
    Origenes says:

    GPuccio @80

    My understanding of neutral theory’s central claim is that (near) neutral mutations can take a simple primitive system X with function A to a complex system Y with function A.

    IOWs starting with a simple primitive system X, it is assumed that there is ‘a ladder of simple gradual naturally selectable neutral variations’ leading up to a complex system Y, which contains CSFI.
    Note the shift from “naturally selectable” (neo-darwinism) to “neutral”.

    Would you say that my understanding is correct?

  92. 92
    gpuccio says:

    Origenes:

    Yes, your understanding is perfectly correct.

    The problem is that, with neutral variation, the probabilities of getting function are exactly the probabilities of a random walk, because it is a random walk and nothing else.

    Genetic drift, as I have said many times, does not change anything, because it is random: any mutation has the same probabilities as any other to be fixed.

    So, a purely neutralist intepretation is exactly the same as saying that, say, ATPsynthase beta chain came into existence by simple change. IOWs pure magic,

    That’s why, when confronted with the probabilistic problem, neutralists always recur to the “help” of NS! 🙂

  93. 93
    Origenes says:

    GPuccio @92

    GPuccio: The problem [with neutral theory] is that, with neutral variation, the probabilities of getting function are exactly the probabilities of a random walk, because it is a random walk and nothing else.

    Indeed, but to be fair, this is an improvement on the probabilities faced by adaptionist neo-darwinism. That is, in as far as neutral theory is able to steer away from natural selection. As we both agree, with Larry Moran, natural selection generally acts as a break on evolution as a blind search. IOWs, contrary to Dawkins’ conviction, evolution driven by natural selection performs worse than a random walk (a blind search).
    So, again, it seems that neutral theory attempts to improve on that. However, an unassisted search, a random walk, is a modest improvement:

    wiki:

    Dawkins: I don’t know who it was first pointed out that, given enough time, a monkey bashing away at random on a typewriter could produce all the works of Shakespeare. The operative phrase is, of course, given enough time. Let us limit the task facing our monkey somewhat. Suppose that he has to produce, not the complete works of Shakespeare but just the short sentence ‘Methinks it is like a weasel’, and we shall make it relatively easy by giving him a typewriter with a restricted keyboard, one with just the 26 (capital) letters, and a space bar. How long will he take to write this one little sentence?

    The scenario is staged to produce a string of gibberish letters, assuming that the selection of each letter in a sequence of 28 characters will be random. The number of possible combinations in this random sequence is 27^28, or about 10^40, so the probability that the monkey will produce a given sequence is extremely low. Any particular sequence of 28 characters could be selected as a “target” phrase, all equally as improbable as Dawkins’s chosen target, “METHINKS IT IS LIKE A WEASEL”.

    Unassisted search will not work for even moderately sized problems [1, 2]. Monkeys at a typewriter eventually generating the complete works of Shakespeare is simply impossible. Simple combinatorics show the universe as modeled by science today is not sufficiently old nor big enough to support the search for even a hand full of pages [3, 4]. Neither are the 10^1000 parallel universes hypothesized by M-theory [1].
    [Dembski]

  94. 94
    DATCG says:

    Gpuccio @92 and rest of discussion by all above. Enjoying this post. Thanks for your contributions again here at UD.

    You said,
    “The problem is that, with neutral variation, the probabilities of getting function are exactly the porbabilities of a random walk, because it isa random walk and nothing else.”

    Question, does Neutral Theory solve the problems Darwinist must overcome for a blind, unguided process of random mutations and Natural Selection?

    Neutral Theory by Kimura was an attempt to address real world problems that Haldane appropriately identified at the time.

    An attempt to get around the cost of substitution and reproductive rates for mammals. For example between chimps and humans.

    Standard evolutionary model(RM & NS) ran out of time. There was not enough time and generations for enough variation and positive selection to take place. Kimura’s solution, Neutral Theory would supposedly allow mutations to build up faster due to duplication errors.

    Neutral Model gave rise to “JUNK” DNA. Yet another dilemma to overcome by Darwinist.

    It seems in trying to solve one real world problem(Haldane’s Dilemma and Kimura’s own admission it was a problem); Neutral theory created more problems and riddles to solve.

    Especially since ENCODE and rise of epigenetics showing more function and less “junk” each day.

    I guess it was a valiant attempt to solve a real world problem of time theoretically. But did it?

    Thanks again. Been busy, but always enjoy your post and discussions that follow.

  95. 95
    DATCG says:

    Orignes @93 ha! I was typing my comment and did not see yours.

  96. 96
    gpuccio says:

    DATCG:

    “Question, does Neutral Theory solve the problems Darwinist must overcome for a blind, unguided process of random mutations and Natural Selection?”

    Absolutely not!

    Origenes is right. Dembski is right. To rely only on random walks to explain proteins, or any other form of complex functional information in biology, is utter folly!

    Only a design perspective makes sense! 🙂

  97. 97
    DATCG says:

    ET @89,

    on “squeezing” found this info online from a Biology 330 level course. Highlighting in block quote(see #2)…

    F-class Pump: functions to power the synthesis of ATP from ADP and Pi by movement of protons from the exoplasmic to the cytosolic face of the membrane down the proton electrochemical gradient.
    ? Rotation of protein shaft couples proton flux.
    ? Alpha and beta subunits arranged alternatively around a hydrophobic shaft that is formed by the Y subunit.
    ? Each alpha/beta subunit has an adenine nucleotide binding site. As the shaft rotates, e- are accepted and this acceptance allows the conformational change for the shaft to rotate.

    Each of the 3 beta subunits differ in conformation and affinity for nucleotides.

    1. Beta unit accepts ADP and Pi and rotates into loose state .
    2. Turns against into tight state where ADP and Pi are squeezed into ATP.
    3. Turns again to open state and releases ATP.

    Energy provided by the electrochemical gradient of H+ is required for the Y subunit to drive each successive transition from loose to tight. ATP is released every 3rd turn.

    After ATP is released, H+ gets pumped back into the matrix by the ATP pump then combine with oxygen to make water. This is critical because, without oxygen, they will accumulate and the concentration gradient needed to run the ATP pumps will not allow the pumps to work.

    Proton motive force also powers the ATP/ADP antiporter.
    Why is this process important?

    Because force is needed to get used ADP into the matrix and newly made ATP to the rest of the cell. They both go through an antiporter to do so. Moving in/out of the membrane.

    Update link:
    ADP and Pi squeezing

  98. 98
    gpuccio says:

    DATCG:

    Yes, it’s a wonderful machine indeed! 🙂

    It’s certanly not a case that the two units that form the body of the F1 part (responsible for the squeezing and the rest) are specially conserved, therefore perfect examples of complex functional information.

    Homology between E. coli and humans:

    Alpha chain (553 AAs):

    290 identities; 373 positives; 561 bits

    Beta chain (529 AAs):

    334 identities; 383 positives; 663 bits

    More than 1200 bits of functional information which can be dated almost at the beginning of biological history, and have remained the same through billions of years!

  99. 99
    Corey Delvine says:

    Ok, I don’t have the time to argue each one of your points, especially if you are going to only quote half my arguments and then “refute” (or belittle) them.

    What gpuccio asked for was “some reasonable argument if why negative selection act in favor of evolution, or is netral, or whatever.”
    None of the quotes you guys give even come close to saying “negative selection works against evolution,” save for Moran’s quote… and we all know how he likes to say things off the cuff.
    All other quotes and pretty much everything I’ve seen point toward negative selection being neutral to the process of evolution as a whole.

    Gpuccio, I think this quote demonstrates your misunderstanding of evolution the best:
    “If you want to argue that such a preservation is the premise for further evolution, I agree. But that does not mean that negative selection has any active role in promoting evolution”
    So preserving what is “fit enough” for the environment is a premise for further evolution, and you agree.
    But you don’t think that this preservation promotes evolution.
    I would argue that holding a population in a state of relatively high fitness is essential to evolution and primes the population for further evolution.

    Let’s think about it this way (it’s a simplification, but it may help):
    Picture a histogram of a single generation, of a single population with fitness on the x-axis (left is less-fit, right is more-fit), the y-axis is the number of individuals with that fitness.
    Now, what does this plot look like?
    I think it resembles a bell curve, do you?
    What will happen if we plot the next generation?
    The curve will probably be ever so slightly shifted to the right, right?
    Now, do you think there is a steep drop-off on the left side of the curve where a certain low-level fitness is culled from the generation?
    Or do you think there is a gentle slope and then eventually a short drop?
    I think it’s a gentle slope on the left side for each generation.
    The left side of this curve is where things like your “substitutions A and B” are free to occur.

    The problem is that first you say “A and B, individually, are slightly deleterious”
    Then you switch to “if the deleterious effect of each individual mutation is strong enough to trigger negative selection”
    You can’t have both.
    Either they’re slightly deleterious, and they remain in the population (in individuals on the left half of the curve fro above)
    Or they’re deleterious enough to trigger negative selection and there are few or no individuals with that mutation within the curve.

    Now the question becomes this:
    How prevalent are slightly deleterious mutations versus mutations that immediately trigger negative selection.

    You seem to think that mutations immediately triggering negative selection is the rule, while slighlty deleterious is the exception.
    That is not the case.
    Now I’m sure your next move will be to cite the thousands of papers that mutate a single nucleotide and results in the death of the organism or some serios defect because I know you are semi-well-read in biology.
    But once again, it’s an issue of perspective.
    What if I told you that publications were extremely biased toward these types of mutations?
    I’m not going to go into the details (much to your chagrin apparently), but I am also pretty well-read on the topics at hand and I can tell you that the history of molecular biology, a pretty young field, has largely been dependent on screening for mutations that cause these massive defective changes to an organism.
    The slighlty deleterious mutations are much more difficult to study and much more difficult to draw conclusions from, at least conclusions that are significant according to the almighty p-value.
    This is unfortunate, but it is the current state of research in biology.

    The fact that you can take a population of millions of cells, mutate each one at a different place, and still get the majority of them to grow and divide happily is a testament to the amount of change that is tolerated by biology.

    ET, this will probably be the last time I respond to you.
    “No other enzyme works by force of pressure”
    What you are calling “force of pressure” is actually called a “conformational change” in an enzyme.
    Here are the words from one of the first papers on the mechanism:
    “ATP synthesis results from energy-linked conformational changes that change the binding of reactants at the catalytic site”
    The paper then goes onto say “The suggestions are also applicable to some ATP-linked active transport processes.”
    (Boyer, 1975)
    So in 1975, scientists were already suggesting the existence of other proteins also using this mechanism.

  100. 100
    gpuccio says:

    Corey Delvine:

    Your post #99 is essentially meaningless. I think it deserves at most a couple of very simple corrections:

    1) We agree that negative selection preserves functional information as it exists. But it certainly does not “promote” new change. Indeed, it makes it more difficult.

    So, if when we say “evolution” we mean “the generation of new functional information”, negative selection certainly does not promote that process, but certainly the existence of previous information and its preservation is a premise for it. I cannot be more clear than that, then you can say whatever you like.

    2) I will not even try to understand you ramblings about normal distributions of fitmness, and how they change in youe imagination. I don’t see any reason to do that.

    3) In my argument, that you clearly misunderstand, “slightly deleterious” refers to mutations that can trigger negative selection. If you are aware of how natural selection works in population genetics, you will know that traits are assigned a “selection coefficient”, which determines how NS will act on them.

    Here is what Wikipedia says about it:

    In population genetics, a selection coefficient, usually denoted by the letter s, is a measure of differences in fitness. Selection coefficients are central to the quantitative description of evolution, since fitness differences determine the change in genotype frequencies attributable to selection.

    The following definition of s is commonly used.[1] Suppose that there are two genotypes A and B in a population with relative fitnesses wA and wB respectively. Then, choosing genotype A as our point of reference, we have wA=1, and wB=1+s, where s measures the fitness advantage (positive s) or disadvantage (negative s) of B.

    For example, the lactose-tolerant allele spread from very low frequencies to high frequencies in less than 9000 years since farming with an estimated selection coefficient of 0.09-0.19 for a Scandinavian population. Though this selection coefficient might seem like a very small number, over evolutionary time, the favored alleles accumulate in the population and become more and more common, potentially reaching fixation.

    So, as you can see, it is not at all necessary that a deleterious mutation be incompatible with life to trigger negative selection. it is not even necessary that it is “strongly deleterious”.

    IOWs, it is enouth that the selection coefficient is negative, slightly negative (but not insignificantly negative) for negative selection to act.

    So, why do you say:

    The problem is that first you say “A and B, individually, are slightly deleterious”
    Then you switch to “if the deleterious effect of each individual mutation is strong enough to trigger negative selection”
    You can’t have both.

    Of course I can have both! Many mutations are “slightly deleterious”, and are acted against by negative selection.

    Your bad faith is proven in the following statements, when you say:

    You seem to think that mutations immediately triggering negative selection is the rule, while slighlty deleterious is the exception.
    That is not the case.

    (emphasis mine)

    First of all, I never used the word “immediately”. You make it up. My statement, quoted by you, was:

    “But if the deleterious effect of each individual mutation is strong enough to trigger negative selection, then the probability of having both mutations in a certain evolutionary time in the same individual becomes essentially zero, because whatever the first mutation is that takes place (A or B), it will be eliminated by negative selection.”

    Where is the “immediately”?

    We are discussing evolutionary times here. There is no need at all for anything to be “immediate”.

    If a mutation is so deleterious that it is incompatible with life, then the individual will just not survive.

    But if it is only “slightly deleterious”, then negative selection will act against it in evolutionary times, according to its selection coefficient. The probability of the trait to survive in its mutated form will be lower than for a neutral mutation.

    Second, I have not discussed at all “rules” or “exceptions” not have I said how frequent IMO each type of mutation is.

    The argument is highly controversial even among darwinists.

    Just for you convenience, I will now say what I think, but it’s just my idea.

    I think that many mutations, probably most mutations, are neutral or quasi-neutral (IOWs minimally deleterious, but not enough so that they behave differently from neutral mutations.

    I think that many mutations, probably a big minority, are slightly deleterious: enough so that negative selection acts probabilsitcally against them.

    I think that a few mutations are strongly deleterious, up to being incompatible with life. Those are strongly selected against by negative selection. Maybe not immediately, but in shorter times.

    I think that almost no mutation is “beneficial”. The few that are, are really an exception. However, those few exception can be positively selected, especially in contexts where a veri strong environmental pressure (against the previous trait) is present, like the antibiotic resistance scenario.

    However, all those “beneficial” mutations are simple enough, and can at most tweak a little some already existing function. They can never generate complex functional information.

    That’s all.

  101. 101
    ET says:

    LoL!@ Corey- Much more has been learned about ATP synthase since 1975. But go ahead in live in the past. It doesn’t matter as you still don’t have a mechanism capable of producing ATP synthase.

    Heck with mutations and natural selection, negative selection and positive selection. Yours doesn’t have a mechanism to account for the existence of living organisms. Even given replicating molecules Spiegelman’s Monster is an obstacle you cannot surmount. Hence yours has to be given starting populations of prokaryotes that already have many irreducibly complex systems up and running.

    So in 1975, scientists were already suggesting the existence of other proteins also using this mechanism.

    And yet they haven’t found any similar protein machines. Nor would they have a mechanism capable of creating it if they did find them.

    “ATP synthesis results from energy-linked conformational changes that change the binding of reactants at the catalytic site”

    That conformational change occurs via the squeezing action.

  102. 102
    Corey Delvine says:

    When I say “evolution”, I mean “making populations more fit to their environment”

    You, however, define evolution using words like “functional” and “information” because you have an agenda.

    “I will not even try to understand you ramblings about normal distributions of fitmness”
    So you don’t want to actually think, ok got it.
    Or do you know that actually thinking about evolution in a sensible way reveals all the faults in your logic and your misunderstandings about biology?

    I’m glad you realize that we are talking about evolutionary timescales.
    Now maybe you realize the fault in your mutations argument:
    Allowing these slightly deleterious mutations to hang around for any time at all (since we are talking evolutionary timescales), it is long enough to account for the low probability of both mutations occurring in the same individual.

    Just for your convenience, I will now say what I think.
    You are the poster child for confirmation bias.
    You make claims about what evolution is incapable of and point to a protein that was fixed in all organisms billions of years ago as support for your argument.
    That’s like trying to convince someone that planes don’t exist by bringing them outside and pointing to the cars on the street.

    ET, you are absolutely impervious to new information, just like Gpuccio here.
    Although, Gpuccio is at least smart enough to make the conscious decision to do so.

    That’s all.

  103. 103
    gpuccio says:

    To all:

    This OP was born as an “instant OP”, in a sense, with the purpose of translating a veri interesting discussion that was taking place in an older thread, and which had in some way focused on important issues, mainly related to the role of NS.

    I am very satisfied of the discussion here, because I believe that a lot of relevant aspects have been touched.

    Gordon Davisson has offered many good arguments, and I have addressed them from my point of view.

    Corey Delvine has offered some arguments, and I have addressed them from my point of view.

    That’s fine.

    But… There is a but.

    I realize that, in the course of the discussion, I have repeatedly tried to focus on one point, which is IMO the strongest, most important argument against the role that NS is supposed to have in the neo-darwinian scenario: the simple fact that complex functions cannot be deconstructed into simple naturally selectable steps.

    Now, my impression is that, in spite of my efforts to emphasize that point, it has not been addressed at all by my interlocutors “on the other side”.

    So, at the cost of being repetitive, I would like to state again that single point here, and to follow it with a challenge.

    So, the statement first. I will do it in two different ways.

    First, I will paste here the summary of the argument that I have already posted at the end of #53. I quote it here entirely, keeping the original emphasis:

    But, in the end, the strongest argument against the power of positive NS remains, IMO, the one that I have tried to emphasize throughout this thread:

    a) Positive Natural Selection can only act on what exists, and the supposedly beneficial variations that arise by chance are always informationally simple.

    b) New complex functional information cannot be deconstructed into simpler steps, each of them more functional and naturally selectable. For NS to have a chance with complex information, such a ladder of simple gradual selectable variations should exist for each example of complex functional information, IOWs for each existing protein. As far as we know, it exists for none.

    And there is no reason at all for such a ladder to exist. It does not exist for language, it does not exist for software, it does not exist for proteins.

    It does not exist for any functional unit where the function depends critically on the final configuration of many bits and is not present unless that functional configuration has been completely implemented.

    This is, IMO, the true and final argument against the power of NS to build up complex functional information. It simply can’t.

    OK, that’s my way of saying it.

    Now, just to give another flavour of the argument, I will paste here the brilliant summary that Origenes has made of it at his post #63 (I hope he does not mind!). Here it is:

    GPuccio @61, @53

    Thank you for your clear well-argued explanation.

    So, just for the record:

    (1) For NS to be the cause for complex specified functional information (CSFI), a ‘ladder of simple gradual naturally selectable variations’ leading up to CSFI should exist.
    (2) We find CSFI in language, software code and DNA code.
    (3) There exists no ‘ladder of simple gradual naturally selectable variations’ for CSFI in language and software code.
    (4) We have not found a ‘ladder of simple gradual naturally selectable variations’ for CSFI in DNA code.

    Therefore (from 3 and 4)

    (c) We have no reason to believe that NS is the cause for CSFI.

    OK, this is the argument.

    Now, the challenge.

    Will anyone on the other side answer the following two simple questions?

    1) Is there any conceptual reason why we should believe that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    2) Is there any evidence from facts that supports the hypothesis that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    Gordon is obviously invited. Corey is invited. Anyone on the other side is invited.

    All those on our side of course, who probably know all too well how to answer, are invited to offer contributions, if they like. 🙂

  104. 104
    gpuccio says:

    Corey Delvine at #102:

    No comment.

    Thank you for the “smart”, anyway…

    By the way, have you read my #103?

    Of course, you are free to answer “no comment” to it! 🙂

  105. 105
    cmow says:

    gpuccio @ 104,
    Also, I think when Corey Delvine said that you were ‘semi well-read’, he was considering that a compliment.

    So, you’ve got that going for you… which is nice.

  106. 106
    Dionisio says:

    gpuccio,

    Sorry to disappoint you, but I think that none of your politely dissenting interlocutors will respond to your challenge @103, because they simply can’t. Anyone who could present a case that meets your challenge should be nominated for the Nobel Prize in biology.

  107. 107
    Mung says:

    I will respond to gpuccio by showing how to evolve an eye in 1% increments!

  108. 108
    ET says:

    Corey:

    When I say “evolution”, I mean “making populations more fit to their environment”

    Too vague to be of any use. More fit can be anything-> faster, slower, taller, shorter, longer, better eyesight, no eyesight.

    Also a more fit X (any organism) is still an X.

    ET, you are absolutely impervious to new information

    You have yet to provide any. On the other hand I have supported my claim. And I also understand why you would focus on the minutiae of how the binding occurs as you sure as heck cannot account for ATP synthase.

    But anyway, there is a paper dealing with two mutations and it doesn’t help you at all:

    waiting for two mutations

    Sheer dumb luck doesn’t seem like a proper scientific mechanism…

  109. 109
    gpuccio says:

    cmow:

    “Also, I think when Corey Delvine said that you were ‘semi well-read’, he was considering that a compliment.

    So, you’ve got that going for you… which is nice.”

    You are right.

    It’s those small rewards that make my life gratifying!

    Well, semi-gratifying, may be… 🙂

  110. 110
    gpuccio says:

    Dionisio:

    “Sorry to disappoint you, but I think that none of your politely dissenting interlocutors will respond to your challenge @103, because they simply can’t. Anyone who could present a case that meets your challenge should be nominated for the Nobel Prize in biology.”

    What’s a challemge, if it’s not a little bit difficult? 🙂

  111. 111
    gpuccio says:

    Mung:

    “I will respond to gpuccio by showing how to evolve an eye in 1% increments!”

    I am a little disappointed! From you, I would have expected at least 0.1% increments!

    Moreover, the evolution of the eye is a rather trendy subject, I would say. Why not try something more original, say the evolution of the ear? I am specially interested in a model for the cochlea and the semicircular canals.

    However, I will try to give you some preciuos help:

    Why not start with a tiny sound sensitive spot? 🙂

  112. 112
    gpuccio says:

    ET:

    Very interesting paper, thank you.

    Some ratehr amazing conclusions by the authors.

    This is for Drosophila:

    From this we see that if both mutations are almost neutral (i.e., relative fitnesses r = (approximately) 1 – 10^-4 and s = (approximately) 1 + 10^-4), then the switch between two transcription factor binding sites can be done in less than1 million years. This is consistent with the results for the even-skipped stripe 2 enhancer mentioned earlier.

    Well, less than 1 million years, in Drosophila, for two mutations that can change the affinitiy of a transcription factor binding site of 10 nucleotides!

    And in humans?

    Our previous work has shown that, in humans, a new transcription factor binding site can be created by a single mutation in an average of 60,000 years, but, as our new results show, a coordinated pair of mutations that first inactivates a binding site and then creates a new one is very unlikely to occur on a reasonable timescale.

    No reasonable time scale? For two functional mutations in a 10 nucleotide site?

    Have they tried to apply that reasoning to the generation of 300+ specific aminoacid sites in the beta chain of ATP synthase?

  113. 113
    Origenes says:

    Corey Delvin: None of the quotes you guys give even come close to saying “negative selection works against evolution,” save for Moran’s quote …

    How do I put this politely? Either Corey has a weird sense of humor or he suffers from a massive comprehension blockage. The quotes provided in #74, #75 and #78 are crystal clear.

  114. 114
    gpuccio says:

    Origenes:

    “The quotes provided in #74, #75 and #78 are crystal clear.”

    I agree.

    However, I think you have put it politely enough!

    But please, be more compassionate: Corey has at least conceded Larry Moran: just to add immediately that Moran is, notoriously, a “naughty boy” in the neo-darwinian field.

    Well, as I usually like naughty boys, I am maybe doubly excited at being, for once, backed up by Larry “naughty boy” Moran! 🙂

    I must say that Dionisio, for example, was not that lucky…

    OK, it’s all about luck, in this life!

    Especially neo-darwinian evolution. 🙂

  115. 115
    gpuccio says:

    To all:

    While I am waiting for answers to my challenge at #103, I would like to make a reflection about this metaphor in Corey Delvine’s last post (#102):

    You make claims about what evolution is incapable of and point to a protein that was fixed in all organisms billions of years ago as support for your argument.
    That’s like trying to convince someone that planes don’t exist by bringing them outside and pointing to the cars on the street.

    (emphasis mine)

    I will not comment on the meaning of the metaphor itself, which frankly eludes me, but rather on its implementation.

    So, here is the reflection:

    Last time I checked, planes and cars were still designed objects.

    So, unless in the meantime planes and cars have started to arise from neo-darwinian evolution and I have not realized it, it seems that Corey is using designed objects in a metaphor, however obscure, about neo-darwinian evolution.

    As this is not at all an isolated case, I wonder:

    Why do neo-darwinists so often resort to designed objects when they want to make examples or metaphors about how neo-darwinian evolution works?

    Is there a reason for that?

    OK, this is not another challenge. I am already busy enough commenting on the many answers to the first one! 🙂

  116. 116
    Dionisio says:

    gpuccio @114:

    I am maybe doubly excited at being, for once, backed up by Larry “naughty boy” Moran!

    I must say that Dionisio, for example, was not that lucky…

    Well, in my case I did deserve to be punished by the distinguished biochemistry professor from the university of Toronto, because I did not ask honest questions.

    For example, I asked:

    “Do you know exactly how the morphogen gradients are formed?”

    Note that I used the ‘tricky’ word ‘exactly’ in a subliminal way, intentionally hidden behind a thin veil or cloud, in order to confuse my interlocutor, who obviously couldn’t notice the presence of such a tricky word.

    An honest question wouldn’t have included such a tricky term at all, or at least it should have highlighted it in bold characters and maybe even underlined:

    “Do you know exactly how the morphogen gradients are formed?”

    Thus the professor would have noticed the ‘tricky’ term within the question and avoided the embarrassingly ridicule answer he gave.

    Mea culpa.

    I should try better next time.

    🙂

  117. 117
    Dionisio says:

    gpuccio @115:

    Why do neo-darwinists so often resort to designed objects when they want to make examples or metaphors about how neo-darwinian evolution works?

    Well, simply because deep in their hearts the know that ‘intelligent design’ is the only valid scientific explanation for the existence of the biological systems. The only show in town.

    That’s all.

    🙂

  118. 118
    Origenes says:

    Are you all gutless and thick or what? Hey atheists you’ve been challenged!

    For your convenience:

    Now, the challenge.

    Will anyone on the other side answer the following two simple questions?

    1) Is there any conceptual reason why we should believe that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    2) Is there any evidence from facts that supports the hypothesis that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    Gordon is obviously invited. Corey is invited. Anyone on the other side is invited.

    [GPuccio #103]

  119. 119
    Mung says:

    Why not start with a tiny sound sensitive spot?

    But that was the hard part, getting from there to an ear is the easy part. You are too easy on me!

  120. 120
    gpuccio says:

    Mung:

    I have a generous heart! 🙂

  121. 121
    Dionisio says:

    Origenes @118:

    Sorry to disappoint you, since your effort to assist is very commendable, but be assured that nobody will take gpuccio’s challenge, because the Neo-Darwinian folks and their comrades lack valid arguments for it. They are loud in vague philosophical discussions that lead nowhere, but they try hard to stay away from serious scientific discussions, specially with gpuccio, who usually gets very technical, though the Italian doctor tries hard to make it easier for the nonscientific folks to understand as much as possible. We have seen it here. The few who dare to engage in discussions end up running for the doors.

    C’est la vie, mon ami.
    🙂

  122. 122
    Dionisio says:

    gpuccio @111:

    Why not try something more original, say the evolution of the ear?

    Ok, that’s easy, look at this:

    https://uncommondescent.com/intelligent-design/the-beautiful-mechanism-by-which-an-egg-becomes-an-embryo/#comment-641041

    🙂

  123. 123
    RodW says:

    gpuccio

    “Do you know exactly how the morphogen gradients are formed?”

    So how did Moran answer this?

  124. 124
    Dionisio says:

    RodW @123:

    How would you answer that question had it been addressed to you?

    Please, note that the original question required just ‘yes’ or ‘no’ for answer.

    The incident took place here in this website around two years ago.

  125. 125
    RodW says:

    Dionisio

    No

  126. 126
    Origenes says:

    Dionisio @124

    Careful please. You better decline RodW’s invitation to derail this thread.
    Why not, instead, accept GPuccio’s challenge Rod?

  127. 127
    gpuccio says:

    Dionisio:

    By the way, have you seen how the Nobel prize was awarded for “discoveries of molecular mechanisms controlling the circadian rhythm”?

    You have a keen intuition in discovering hot issues, when you give links to the literature. 🙂

  128. 128
    gpuccio says:

    RodW:

    These are just a few quotes from Dionisio’s recent activity of literature search:

    How morphogens are transported has been intensively studied.

    In addition to simple diffusion, a number of other transport mechanisms have been proposed and are actively being investigated, including shuttling via other proteins, transport via vesicles (e.g. exosomes and exovesicles) and delivery via cellular projections (e.g. cytonemes).

    According to morphogen gradient theory, extracellular ligands produced from a localized source convey positional information to receiving cells by signaling in a concentration-dependent manner.

    How do morphogens create concentration gradients to establish positional information in developing tissues?

    Surprisingly, the answer to this central question remains largely unknown. During development, a relatively small number of morphogens are reiteratively deployed to ensure normal embryogenesis and organogenesis.

    Thus, the intracellular processing and extracellular transport of morphogens are tightly regulated in a tissue-specific manner.

    Over the past few decades, diverse experimental and theoretical approaches have led to numerous conflicting models for gradient formation.

    it seems clear now that morphogens can be transported in several ways, but it remains unclear to what extent the different routes contribute to patterning.

    A number of studies highlighted the dynamic nature of morphogen read-out and the importance of feedbacks and a network context.

    Important open problems concern the mechanistic basis of scaling and growth control – as well as the role of Turing patterns.

    But, to quote Origenes at #126:

    “Why not, instead, accept GPuccio’s challenge Rod?”

  129. 129
    Dionisio says:

    gpuccio @127:

    Perhaps it has to do with what you wrote @46:

    “See how spending time with neo-darwinists can teach us a lot!”

    They can even teach us to predict the hottest issues in science. Just search for their latest fairytales. How? Look for terms like “co-opted” and things like that. Actually you taught me the meaning of that term.
    Or search their literature for words like “surprisingly”, “unexpectedly” or “strikingly” and look at the papers with higher frequency of those terms.
    Bingo! The method doesn’t fail.
    Maybe that’s what they do at the committee in Stockholm too?

    🙂

    BTW, Did you notice that this thread has received over 1000 visits in about a week? Apparently lots of anonymous readers, onlookers, lurkers attracted to this topic.

    Well done! Congratulations!

  130. 130
    Dionisio says:

    RodW @125:

    The named professor didn’t answer like you.

    BTW, gpuccio graciously provided a tremendous hint @128, but since you responded almost 20 minutes before that comment appeared, your answer did not benefit from it.

  131. 131
    Dionisio says:

    Origenes @126,

    Thanks for the alert. Will keep in in mind.

  132. 132
    Dionisio says:

    gpuccio @128:

    Excellent selection of quotes describing the current status of the given topic. Thanks.

  133. 133
    Dionisio says:

    gpuccio @127:

    By the way, have you seen how the Nobel prize was awarded for “discoveries of molecular mechanisms controlling the circadian rhythm”?

    Doesn’t that open a Pandora’s box and a can of worms that the Neo-Darwinian folks might not be able to close?

    🙂

    BTW, doesn’t this fundamental topic seem very related to what you’ve been saying all along here about the missing complex functional information that must be hidden somewhere in the biological systems?

    Did I get this association wrong?

  134. 134
    Dionisio says:

    @131 error correction:

    Will keep it in mind.

  135. 135
    Corey Delvine says:

    Gpuccio, I’m not attempting to answer your question because it is obvious that you are not willing to have an actual discussion.

    I tried to have a simple discussion with you in comment 99, and you responded with “I will not even try to understand your ramblings”.
    Now, why would you claim to want to have a discussion, but then refuse when actually engaged in one?

    Is it because your idea of a discussion means using your own twisted definitions and examples of proteins that “couldn’t have evolved”?

    No matter what is said in response to your questions, you will end up pointing to ATP synthase and saying “look how complex this is, there’s no way it could have evolved, and you’ll never convince me otherwise!”

  136. 136
    ET says:

    Corey Delvine- The fact that you say:

    “look how complex this is, there’s no way it could have evolved, and you’ll never convince me otherwise!”

    Proves that you don’t want a discussion. It isn’t about mere complexity. It is about functional irreducible complexity. Also it isn’t about “evolution”. It is about evolution by means of blind, mindless processes.

    And yes it is up to the people making the claim that ATP synthase evolved via blind, mindless processes to convince everyone else by using science and not dogma.

    ID has said exactly what will falsify it. IDists have said exactly what will convince us otherwise. And in neither case does our opponents’ whining fit the bill. So good luck with that.

  137. 137
    Mung says:

    We describe two attempts at a quantitative model – one by Sir Fred Hoyle and the other by Dr. Richard Dawkins. While most models modify a previously-existing, functional, complex structure, both of these models are attempts to quantify the ability of evolution to innovate rather than modify. We critique these two models, and present our quantitative model. Specifically, we hypothesize a primitive version of Chlamydomonas reinhardtii, a simple single-celled green alga, which does not yet have a functional eyespot. We then calculate the probability of developing an eyespot that is functional enough to confer an evolutionary benefit assuming all components are pre-existing, except a few proteins.

    The Need For A Quantifiable Model Of Evolution (PDF)

  138. 138
    JVL says:

    ET – But anyway, there is a paper dealing with two mutations and it doesn’t help you at all:

    waiting for two mutations

    Sheer dumb luck doesn’t seem like a proper scientific mechanism…

    Amusingly enough, the authors of that paper point out some of the errors made by Dr Behe. Scroll down to just before the conclusion.

  139. 139
    Origenes says:

    Corey Delvin @135

    The Challenge.

    Will anyone on the other side answer the following two simple questions?

    1) Is there any conceptual reason why we should believe that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    2) Is there any evidence from facts that supports the hypothesis that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    Gordon is obviously invited. Corey is invited. Anyone on the other side is invited.

    [GPuccio]

    Corey Delvin: Gpuccio, I’m not attempting to answer your question because it is obvious that you are not willing to have an actual discussion.

    So, the reason why you won’t answer GPuccio’s question is not because you do not have an answer, but, instead, it is because you do not like GPuccio’s attitude.
    Okay, got it. That is a plausible explanation.

  140. 140
    Dionisio says:

    Origenes @139,

    No, the reason is that they lack what is required in order to engage in a technical discussion with someone of the caliber of our Italian doctor. It’s ‘willingness’+’seriousness’ among other important requirements. That’s all. The rest is pure propaganda disguised as ‘cheap excuses’, just to say it nicely. Can’t fool me. Been there, done that myself. I wasn’t always on this side of the fence. I know how they think on the other side, because I was there quite a long time too.

    Non sono mica nato ieri!

  141. 141
    ET says:

    JVL @ 138- Amusingly enough Dr Behe has responded to that paper and made corrections to their premises.

  142. 142
    gpuccio says:

    JVL:

    “Amusingly enough, the authors of that paper point out some of the errors made by Dr Behe. Scroll down to just before the conclusion.”

    I think we are all aware of that. It’s clearly stated in the abstract.

    In addition, we use these results to expose flaws in some of Michael Behe’s arguments concerning mathematical limits to Darwinian evolution.

    I find it weird, more than amusing. However, what’s the point? Do you think that criticizing Behe chnages in any way their experimental results?

    Which are those that I quoted in #112.

    But, to quote again Origenes:

    “Why not, instead, accept GPuccio’s challenge JVL?”

  143. 143
    gpuccio says:

    Corey Delvine at #135:

    Gpuccio, I’m not attempting to answer your question because it is obvious that you are not willing to have an actual discussion.

    I appreciate that you are the first, up to now, who has ansered my challenge, even if only to state that you will not attempt to answer it.

    It is also comforting to know that the problem is not that you don’t have any answers, but rather that I am not willing to have an actual discussion.

    So, it’s not any lack of arguments on your part, just my personal faults.

    OK, it’s fine with me. I could ask you to give us the answers just the same, for the sake of other onlookers, but I will not do it. After all, sinners cannot beg.

    My personal faults are too serious.

    First, I am not willing to have an actual discussion with you.

    Sure, I have only written 12 comments (including this) addressed to you personally in this thread:

    19
    23
    32
    33
    77
    79
    81
    87
    88
    100
    104
    143

    for a sum total of 35074 characters and 6950 words. A lot of complex functional information, I would say.

    But it seems that I am not willing to have a discussion with you.

    Second, I have called “ramblings” your ramblings at #99. I hope you have not taken offence, but that’s exactly what I think: that they are ramblings, and don’t deserve an answer. This is a form of discussion, too.

    Third, I use “my own twisted definitions”. That’s serious indeed. I confess that I would not be at ease using your own twisted definitions. That’s my limit, I suppose. Human nature.

    Finally, I point to proteins that IMO are too complex to be explained by your theory, and ask you to explain how they could be explained by your theory. Horrible, indeed! What kind of dicsussion is this? I should probably just acknowledge that you believe they can be explained. How arrogant of me to ask for clarifications.

    And, worst of all, I point to specific examples. Like ATP synthase beta chain.

    That’s really unforgivable. Even when you give me beautiful explanations about cars and planes, I remain skeptic, and still mention that ugly protein!

    Even worse, not only that one. In my evil attitude, I have even suggested that you read some of my previous OPs. For example:

    https://uncommondescent.com/intelligent-design/homologies-differences-and-information-jumps/

    https://uncommondescent.com/intelligent-design/information-jumps-again-some-more-facts-and-thoughts-about-prickle-1-and-taxonomically-restricted-genes/

    https://uncommondescent.com/intelligent-design/the-highly-engineered-transition-to-vertebrates-an-example-of-functional-information-analysis/

    https://uncommondescent.com/intelligent-design/the-amazing-level-of-engineering-in-the-transition-to-the-vertebrate-proteome-a-global-analysis/

    https://uncommondescent.com/intelligent-design/interesting-proteins-dna-binding-proteins-satb1-and-satb2/

    where I dare to mention many other proteins with high informational content and informational jumps that need to be explained, like:

    Prickle 1
    Prickle 2
    Astrotactin 1
    Astrotactin 2
    BRNP1
    Cadherin 2
    Integrin alpha-V
    Neural cell adhesion molecule 1
    SATB1
    SATB2

    and even make a computation of the information jump in vertebrates for the whole human proteome, giving amazing numbers like 1.7 million bits!

    You have all the reasons in the world not to take my challenge, indeed. 🙂

  144. 144
    Corey Delvine says:

    Ok gpuccio I have skimmed some of what you have pointed me to , so how about a variation of your challenge:

    Pick your single best example of an informational jump in a protein.

    Tell us what the jump in bits is.

    Tell us the two organisms you use to calculate this jump.

    Tell us the number of years ago that the last common ancestor of these two organisms existed.

    Tell us the number of years ago that each of these two organisms appeared.

    Again, all we need is three names, a number of bits, and three dates, and we will have a discussion about that.

  145. 145
    RodW says:

    Dionisio @130
    gpuccio @128

    The way I’d explain it is: a gradient can be produced if a morphogen is made at one side and degraded at some distant site.
    Another way is for the morphogen to be produced for some limited period of time and then the gradient stabilized

  146. 146
    Origenes says:

    Corey Delvin @99

    Corey Delvin: Gpuccio, I think this quote demonstrates your misunderstanding of evolution the best:

    “If you want to argue that such a preservation is the premise for further evolution, I agree. But that does not mean that negative selection has any active role in promoting evolution.” [GPuccio]

    So preserving what is “fit enough” for the environment is a premise for further evolution, and you agree.
    But you don’t think that this preservation promotes evolution.

    Well of course GPuccio does not think that preservation has an active role in promoting evolution. Why would anyone? “Preservation”, by definition, means that organisms do not evolve/change. For heaven’s sake, that’s what the word means!
    In fact, it can be rightfully said that, when the original sequence is preserved by weeding out new mutations, preservation acts as a break on evolution/change. What is unclear about that?

    Corey Delvin: I would argue that holding a population in a state of relatively high fitness is essential to evolution and primes the population for further evolution.

    And how is that different from what GPuccio says? Again, GPuccio: “If you want to argue that such a preservation is the premise for further evolution, I agree.”
    So, why is it that you claim that the quote shows GPuccio’s alleged misunderstanding of evolution? What is your point?

  147. 147
    Dionisio says:

    gpuccio @143,

    Doctor, you’re a master of patience and politeness.

    I would have ignored your interlocutor long ago.

    Maybe that’s one of the various reasons why I’m not allowed to moderate any discussion thread? 🙂

  148. 148
    ET says:

    Cory:

    Pick your single best example of an informational jump in a protein.

    Can your position even account for the existence of proteins? I doubt it so perhaps you should start with that.

  149. 149
    Eugene S says:

    GP,

    What a great effort! Thank you very much. I have read the OP now, which is an achievement in its own right 🙂 I am still to get through the Q&A thread, which I will do in due course as soon as time allows. In particular, I will read your response to my question about gene duplication.

    I always enjoy reading your posts.

    I have one comment so far. As I started reading your OP, it crossed my mind and then I noticed GD mentioning “recombination”, and said to myself: “A-ha!”

    As a matter of fact, in terms of search algorithm capabilities, I suspect recombination will improve the chances! I am not a biologist, as you already know, but I have spent some time doing combinatorial search in operations research and scheduling. There is a simple standard local search, whereby you start from an initial schedule (a functional protein in this context) and tweak it by removing one job (AA in this context) at a time. The neighbourhood of a solution (i.e. what modifications you can make to the initial schedule) is determined by all the possible individual job deletions/reinsertions into the initial schedule.

    However, there is also what is known as large neighbourhood search, whereby you at a single search iteration retract a group of related jobs (where ‘relatedness’ is determined in some clever way so as to improve the capabilities of the search). The point is that it can considerably increase the neighbourhood your local search is exploring. Consequently, it can have a tangible impact on what the search can achieve probabilistically, because a modified schedule/protein that is unreachable by conventional (Darwinian?) local search can now become reachable.

    Now, having said this, I understand that recombination will probably not qualify as neo-Darwinian. I don’t know what such recombinations would involve, in biological terms. Nonetheless, it would be really interesting to hear your further comments on this.

    The original implementation of Large Neighbourhood Search as a meta-heuristic (i.e. as something that helps the search get out of local optima) was proposed by Paul Shaw. Actually, I know the guy. I met him when I worked as a researcher for University of Glasgow. 🙂

    P. Shaw. Using constraint programming and local search methods to solve vehicle routing problems, Fourth International Conference on Principles and Practice of Constraint Programming, v 1520, Lecture Notes in Computer Science, pp 417–431, 1998.

  150. 150
    gpuccio says:

    Corey Delvine at #144:

    OK, perfect! Let’s do serious work.

    You ask:

    Ok gpuccio I have skimmed some of what you have pointed me to , so how about a variation of your challenge:

    Pick your single best example of an informational jump in a protein.

    Tell us what the jump in bits is.

    Tell us the two organisms you use to calculate this jump.

    Tell us the number of years ago that the last common ancestor of these two organisms existed.

    Tell us the number of years ago that each of these two organisms appeared.

    Again, all we need is three names, a number of bits, and three dates, and we will have a discussion about that.

    That’s easy. All the proteins I mentioned in my post #143 are good examples of informational jumps that are extremely problematic for the neo-darwinian paradigm. What they have in common is that the jump can be demonstrated in the first vertebrates, which is the issue I have focused upon in my last OPs.

    If I have to choose one, let’s it be SATB1, which is the one I analyzed more recently, together with its cousin protein SATB2.

    You can find all the details here:

    https://uncommondescent.com/intelligent-design/interesting-proteins-dna-binding-proteins-satb1-and-satb2/

    But I will summarize things here, for your convenience:

    Protein: SATB1
    Length of the human form: 763 AAs
    Accession number (Uniprot): Q01826

    The jump is measured as the difference in human-conserved homology as measured in two groups:

    a) All Metazoa that are not vertebrates (from now on: pre-vertebrates)

    b) Cartilaginous fish

    Jump in functional bits (measured as human-conserved BLAST homology) between pre-vertebrates and vertebrates:

    1049 bits

    The two organisms where the jump is calculated are:

    1) Parasteatoda tepidariorum (a spider)

    homology with the human protein:

    154 bits (78 identities, 111 positives, E= 3^-37)

    This is the best hit among all non vertebrate animal. In this case spiders have better hits than deuterostomia, who have even lower homology, for example:

    Saccoglossus kowalevskii (hemichordates): 87 bits
    Branchiostoma floridae (cephalochordates): 69.3 bits

    2) Rhincodon typus (Whale shark)

    homology with the human protein:

    1203 bits (603 identities, 659 positives, E= 0.0)

    This is the best hit in cartilaginous fish. Callorhinchus milii has very similar values (1184 bits).

    So, the jump in vertebrates is the difference between the two hits:

    1203 – 154 = 1049

    The functionality of the new added information is demonstrated by its conservation for 400+ million years (from cartilaginous fish to humans).

    The evolutionary time when this new functional information appeared is in the window between the appearance of the first chordates and the appearance of the first vertebrates, IOWs between the common precursor of chordates and the common precursor of vertebrates, before the split between cartilaginous fish and bony fish (because humans derive from bony fish).

    While no one can give an exact time window, a reasonable guess is between 440 and 410 million years ago, where:

    440 million years ago is a reasonable guess for the split between deuterostomia (including chordates like cephalocordata and Tunicates) and the first Vertebrates

    410 million years ago is a reasonable guess for the split between cartilaginous and bony fish

    Therefore, the approximate window of evolutionary time for the generation of the new functional information is: about 30 million years.

    I hope that answers your questions. Let the discussion begin.

  151. 151
    Dionisio says:

    Here’s a summary of our mini-exchange on the Canadian biochemistry professor:
    RodW @123:

    “Do you know exactly how the morphogen gradients are formed?”

    So how did Moran answer this?

    D @124:

    RodW @123:

    How would you answer that question had it been addressed to you?

    Please, note that the original question required just ‘yes’ or ‘no’ for answer.

    The incident took place here in this website around two years ago.

    RodW @125:

    No

    D @130:

    RodW @125:

    The named professor didn’t answer like you.

    BTW, gpuccio graciously provided a tremendous hint @128, but since you responded almost 20 minutes before that comment appeared, your answer did not benefit from it.

    RodW @145,

    The way I’d explain it is: a gradient can be produced if a morphogen is made at one side and degraded at some distant site.
    Another way is for the morphogen to be produced for some limited period of time and then the gradient stabilized

    RodW, your answer @125 was correct. The question you quoted @123 required only ‘yes’ or ‘no’ for answer. Nothing else should be added.

    Based on what gpuccio quoted @128, what you wrote @145 practically doesn’t explain anything. What were you trying to comment on?

  152. 152
    RodW says:

    Dionisio

    Weren’t you asking me how I would explain morphogen gradients?

  153. 153
    gpuccio says:

    Eugene S:

    Your post #149 is very interesting, but I would like to clarify a few point, and then hear what you think.

    When we consider the probabilities to find some functional target by a random walk, if our scenario is a new functional isalnd, IOWs a functional sequence which is essentially unrelated to the starting sequence, whatever it is, then our serach is not a neighbourhood search.

    Indeed, any unrelated state will be as likely as any other.

    IOWs, we are in the ocean of the search space, and we have absolutely no reference which can guide the search.

    In biology, neighbourhood searchs can be useful only in small tweakings of an existing function, and in that case negative selection can anyway be a problem, as we have discussed here in great detail.

    However, the real problem for neo-darwinism, and the real point of ID, is generating new complex functional information, like a new superfamily, then the search really has to travel the ocean of the search space. Which, as you know, is really huge.

    Indeed, when you start from random sequence libraries, you are already testing isolated points in the search space.

    The 2000 basic protein superfamilies are completely unrelated at sequence level (and at structure and function level too), so that a transition from one to anotger is certainly not a neighborhood search.

    If a protein acquires 1000+ bits of information in the transition to vertebrates, that is not a neighbourhood search.

    Recombination has a supposed role in remixing existing information in new combinations, like in exon shuffling.

    Random variation has many ways to move directly into the ocean of the search space, whatever the starting sequence. For example, any frameshift mutation will change completely the sequence to be translated.

    Indels, transposon activity, translocations, and so on can change and remix existing sequences. But they do that in a random way. So any state has the same probabilities to be reached.

    Non coding sequences which become coding sequences (a scenario ever more likely for protein origin) are another good example of jumps into the full ocean of the search space, because the sequence in non coding regions has no relationships of necessity with its possible meaning if transcribed and translated. Agian, we are in the field of mere chance.

    Unless, of course, sequences are designed to become protein coding genes. 🙂

  154. 154
    gpuccio says:

    RodW:

    I think that some interesting problems about morphogen gradients could be:

    Is the gradient generated only by synthesis, diffusion and degradation? Is really diffusion the main factor here? What is the role of other transport systems, for example cytonemes? How is the gradient controlled and read? What is the risk of error, and what are the systems of error correction?

    And so on.

    Perhaps you can agree that it is a fascinating field for research.

  155. 155
    Origenes says:

    EugeneS @149

    EugeneS: As a matter of fact, in terms of search algorithm capabilities, I suspect recombination will improve the chances! I am not a biologist, as you already know, but I have spent some time doing combinatorial search in operations research and scheduling.

    Eugene, you may find these two articles on exon shuffling (recombination) to be informative.

    Excerpt:

    While the hypothesis of exon shuffling does, taken at face value, have some attractive elements, it suffers from a number of problems. For one thing, the model at its core presupposes the prior existence of protein domains. …

    “For those elements to work as robust modules,” explains Axe, “their structure would have to be effectively context-independent, allowing them to be combined in any number of ways to form new folds.” In the case of protein secondary structure, however, this requirement is not met. ….

    A further issue relates to interface compatibility. The domain shuffling hypothesis in many cases requires the formation of new binding interfaces. Since amino acids that comprise polypeptide chains are distinguished from one another by the specificity of their side-chains, however, the binding interfaces that allow units of secondary structure (i.e. ?-helices and ?-strands) to come together to form elements of tertiary structure is dependent upon the specific sequence of amino acids. That is to say, it is non-generic in the sense that it is strictly dependent upon the particulars of the components. Domains that must bind and interact with one another can’t simply be pieced together like LEGO bricks. ….

    In fact, in the few cases where protein chimeras do possess detectable function, it only works for the precise reason that the researchers used an algorithm (developed by Meyer et al., 2006) to carefully select the sections of a protein structure that possess the fewest side-chain interactions with the rest of the fold, and chose parent proteins with relatively high sequence identity (Voigt et al., 2002). This only serves to underscore the problem. Even in the Voigt study, the success rate was quite low, even with highly favorable circumstances, with only one in five chimeras possessing discernible functionality.

  156. 156
    Mung says:

    Eugene S

    As a matter of fact, in terms of search algorithm capabilities, I suspect recombination will improve the chances!

    https://mitpress.mit.edu/books/compositional-evolution

  157. 157
    Dionisio says:

    RodW @152:

    Dionisio

    Weren’t you asking me how I would explain morphogen gradients?

    No. You may want to read carefully what I wrote to you, which is quoted @151 for your convenience.

    It would not make much sense for me to ask somebody to explain morphogen gradient formation or interpretation. Those are huge topics in and by themselves.

    I could ask someone to reference some serious literature on the subject, though.

  158. 158
    Corey Delvine says:

    Hmm…30 million years, that’s a long time.
    35 bits per million years then?
    Does that sound fast to you?
    Is there any information on typical rates for other proteins in various organisms with varying gaps in time, or have you calculated any of these yourself?

    Also, where are you getting 440mya and 410mya from?
    We are comparing the human sequence to the whale shark and a spider.
    The difference in time that spiders diverged from the human lineage and the time that whale sharks diverged from the human lineage is the time period when this jump in bits took place, correct?
    Do the 440 and 410mya represent these times, repsectively?

  159. 159
    ET says:

    Corey:

    The difference in time that spiders diverged from the human lineage and the time that whale sharks diverged from the human lineage is the time period when this jump in bits took place, correct?

    Question begging

  160. 160
    gpuccio says:

    Corey Delvine:

    You say:

    Hmm…30 million years, that’s a long time.
    35 bits per million years then?

    I suppose you are dividing 1049 bits by 30 million years, aren’t you?

    1049 / 30 = 34.96666667

    OK, I will try to put it as friendly as possible:

    This is simply a gross mistake you are making.

    The “bits” we are considering here are a measure of improbability/information, but they ar not a linear measure. They are exponential.

    Therefore, 1049 bits means:

    a probability of 1:2^1049

    That number measures the probability of getting that sequence homology by chance.

    2^1049 = 10^315

    And, just to be clear:

    2^1049 / 30 = 2^1049 / 2^4.9 = 2^1044

    IOWs, 1044 bits, not 35 bits.

    Now, as a little bit of irony in a serious discussion is usually fine, let me quote your previous statement, slightly modified:

    If you miss the simple fact that measures of information are exponential in base two, then:

    you just don’t understand how ID theory works! 🙂

    (Dionisio, that sounds good, doesn’t it?) 🙂

    Does that sound fast to you?

    What do you say?

    Finding a sequence that has 1 probability out of a search space of 2^1049 (10^315) of being found by a random search in a time of 30 million years seems not only fast, but beyond conception.

    You cannot find such a target in 30 million years, not in 100 million years, not in 5 billion years. Not by a random search.

    Unless, of course, you can deconstruct that sequence into many simpler steps, each of them naturally selectable. that would allow the intervention of NS, and could lower the probabilistic barriers.

    Which brings us back to my challenge!

    Then you say:

    Is there any information on typical rates for other proteins in various organisms with varying gaps in time, or have you calculated any of these yourself?

    I am not sure what you mean here. However, I suppose you could be interested in what is said in the paper “Waiting for two mutations” (yes, the one where the authors criticize Behe, and the one that has already been debated in some detail here). I quote what I said at #112:

    Some rather amazing conclusions by the authors.

    This is for Drosophila:

    From this we see that if both mutations are almost neutral (i.e., relative fitnesses r = (approximately) 1 – 10^-4 and s = (approximately) 1 + 10^-4), then the switch between two transcription factor binding sites can be done in less than1 million years. This is consistent with the results for the even-skipped stripe 2 enhancer mentioned earlier.

    Well, less than 1 million years, in Drosophila, for two mutations that can change the affinitiy of a transcription factor binding site of 10 nucleotides!

    And in humans?

    Our previous work has shown that, in humans, a new transcription factor binding site can be created by a single mutation in an average of 60,000 years, but, as our new results show, a coordinated pair of mutations that first inactivates a binding site and then creates a new one is very unlikely to occur on a reasonable timescale.

    No reasonable time scale? For two functional mutations in a 10 nucleotide site?

    Two specific nucleotide mutations in a 10 nucleotide sequence are a rather simple event. Two specific nucleotides bear only 4 bits of information at most, and a whole sequence of 10 specific nucleotides bears at most 20 bits of information, because the search space here is 4^10.

    And they are talking of “less that 1 million years” for Drosophila!

    We were discussing 1049 bits!

    And remember, these numbers are exponential measures!

    I will answer your last question in next post.

  161. 161
    gpuccio says:

    Corey Delvine:

    You ask:

    Also, where are you getting 440mya and 410mya from?
    We are comparing the human sequence to the whale shark and a spider.
    The difference in time that spiders diverged from the human lineage and the time that whale sharks diverged from the human lineage is the time period when this jump in bits took place, correct?
    Do the 440 and 410mya represent these times, repsectively?

    These are perfectly legitimate questions about my methodology. I will gladly answer them.

    Yes, those number represent the time window when the appearance of the new information took place. Of course, as I have said explicitly, they are “educated guesses”, because we don’t have absolutely certain estimations for those tiimes. I have tried to get the best estimation from what I have found all around.

    But what are the times we are speaking of?

    OK, 410 million years is my estimation for “the time that whale sharks diverged from the human lineage”, as you correctly say. More in general, for the time of the split between cartilaginous fish and bony fish (which are the ancestors of humans).

    The second time, 440 million years, is, similarly, the time that pre-vertebrates diverged from the human lineage. That corresponds to the time of the split between non vertebrate chordates (IOWs, Cephalochordata and Tunicata) diverged from vertebrate chordates (IOWs, the common ancestor of cartilaginous fish and bony fish).

    The relevant splits here are:

    Protostomes-deuterostomes

    Other Deuterostomes (echinoderns, emichordates)-Chordates

    Non vertebrate chordates (Cefalochordata, Tunicata) – Vertebrates

    The last split is the one we are interested in.

    My methodology, however, is to take the highest hit in the whole group of pre-vertebrates and the highest hit in cartilaginous fish.

    That is clearly and explicitly stated in MY OPs, and in my post #150 here:

    The jump is measured as the difference in human-conserved homology as measured in two groups:

    a) All Metazoa that are not vertebrates (from now on: pre-vertebrates)

    b) Cartilaginous fish

    The reason for that is very simple: I want to get the highest detectable homology to the human form of the protein before the appearance of vertebrates, and compare that to the highest detectable homology to theb human for of the protein before the appearance of bony fish.

    That ensures an unbiased evaluation of the informational jump, because any homology that emerged before the final jump is considered as already available information, whenever it emerged, and is subtracted to the homology in cartilaginous fish, to get a measure of the informational jump.

    Now, in general we expect to find the highest homology to humans in Cephalochordata or Tunicata, or at least in deuterostomia. It is often so, but not always.

    Especially when homologies are low, we can find some higher homology in some protostomes, rather than in deuterostomes.

    Here, for example, the highest hit was with a spider, a protostome. Deuterostomes had lower hits, and that too is clearly and explicitly stated in my post #150:

    1) Parasteatoda tepidariorum (a spider)

    homology with the human protein:

    154 bits (78 identities, 111 positives, E= 3^-37)

    This is the best hit among all non vertebrate animal. In this case spiders have better hits than deuterostomia, who have even lower homology, for example:

    Saccoglossus kowalevskii (hemichordates): 87 bits
    Branchiostoma floridae (cephalochordates): 69.3 bits

    So, my methodology is to use the highest value, wherever we find it in the non vertebrate group.

    The rationale for that is that, if some level of information homologous to the human form can be detected before the appearance of vertebrates, that is the value we must consider and subtract, because that information existed. It could have been still present in the common ancestor before the split of prevertebrates and vertebrates, and then be lost in deuterostomia, and be kept, for some reason, in a spider.

    Of course, there are other possible explanations for the fact that sometimes some protostomia are more similar to humans than deuterostomia, but my methodology ensures that my results are as fair as possible to the darwinist theory, because they consider the highest value of previous information.

    For example, if I had taken the highest hit in Cephalochordata, which are the immediate ancestors of vertebrates, I would have got:

    69.3 bits

    and my result for the jump would have been:

    1203 – 69.3 = 1133.7

    which is higher than the result of 1049 that I get using the value in the spider.

    IOWs, I try to be fair, and to give darwinism the highest chances.

    I hope that is clear.

  162. 162
    Dionisio says:

    gpuccio @160:

    Now, as a little bit of irony in a serious discussion is usually fine, let me quote your previous statement, slightly modified:

    If you miss the simple fact that measures of information are exponential in base two, then:

    you just don’t understand how ID theory works!

    That kind of reminds me of the Star Wars Jedi Counterattack episode. 🙂

    Basically, you served them their own medicine. 🙂

    However, in this case your interlocutor himself justified such a reaction, by missing the exponential character of the important probabilistic parameter associated with the discussed issue.

    Oh, well. What else is new?

  163. 163
    Eugene S says:

    Mung, Origenes

    Thanks for the references. I am aware of models showing that sexual reproduction has an information gain of sqrt(genome size), if I remember correctly. Of course, these are models but I guess they reflect reality at least to some degree.

    That is why I suspect that mere point AA changes are not the only player here as indeed the protein search space is vast. That is to say, if evolution can really do anything tangible in practice, which I do not really expect.

    I still believe that mere recombination cannot get us new complex functionality that is not there at the start, so in general I agree with GPuccio. But I think this needs to be shown more rigorously.

    GPuccio,

    Thanks very much. I am actually jumping over the discussion now, which I still need to catch up with.

    My understanding is that an evomodel has to show that a living system can really traverse large areas of search space where, as we all know, function does not exist. Recombination/large neighbourhood search appear good candidates for it.

    By definition, a local search algorithm starts from a candidate solution (which need not be feasible – that is why I was asking you about gene duplication, where a copy need not be biologically meaningful and can allegedly traverse areas of config space where there is no function) and then iteratively moves to a neighbor solution (which may or may not prove feasible and is of the same, better or worse fitness as the one previously analyzed). In this sense, any local search is a type of neighbourhood search. On the other hand, the more sophisticated the search is, the more active information about the search space it requires and, consequently, the further it departs from the Darwinian model.

    I agree with you that all this literature on genetic algorithms demonstrates the capabilities not of Darwinian search but of intelligent selection. Which is then presented as something allegedly occurring in nature.

    I understand that in cases like large neighbourhood search, in order for the model to work in practice, it needs intelligence because, as you rightly point out, without intelligently controlling neighborhoods (which presupposes some knowledge of problem domain, where solutions are likely to be, the structure of solutions, etc.) we are back to random chance. But I think that we need to show that more rigorously (in terms of probabilistic barriers). That is why I would certainly like to be absolutely clear myself on the issues I raised (such as gene duplication).

  164. 164
    gpuccio says:

    Eugene S:

    I agree with what you say.

    But let’s simplify the problem.

    We have, at some time in evolutionary history, a new sequence that appears for the first time. It has no sequence homology with what was there before, and it is highly specific and functional.

    Let’s take the example of SATB1, which I am currently discussing with Crey Delvine.

    My references are:

    MY OP here:

    https://uncommondescent.com/intelligent-design/interesting-proteins-dna-binding-proteins-satb1-and-satb2/

    And my posts #150, #160 and #161 here.

    OK, please note that when I compute the information jump I am not considering the whole molecule, but only the functional part that appears in vertebrates.

    As we know, that is 1049 bits for SATB1. And, even if we stick only to identities, there are about 525 new identities that appear in vertebrates and are then conserved up to humans.

    Now, the point is, those 1049 bits of functional information, those 525 new identities conserved up to humans, are completely new.

    They have no homology with what existed before, in the whole scenario of metazoa. they appear in vertebrates for the first time.

    Indeed, my methodology measure, the jump in information: only what is new.

    So, the question is: we see a specific sequence of at least 525 specific AAs (without considering the positives, which add to the functional information), which appears in about 30 million years of natural history.

    How?

    Why is it relevant if single point mutations or other forms of recombination acted in those 30 million years?

    Excuse me, this new sequence is new. By definition, it has no homology with what existed before.

    So, no sequence with any “nearness” to this specific functional island existed before vertebrates.

    Remember, “nearness” means nearness in the sequence space. And nearness in the sequence space means detectable homology. No homology, no nearness! 🙂

    So, we have no problems of “near neighbourhood search” or “large neighborhouud search” here. We just have a pure random walk.

    Indeed, near neighborhood search here is penalized, because anyway we know that a large part of the sequence space must be traversed anyway. So, mutations of an existing functional protein are not the best scenario.

    But we can gladly imagine that we start from any possible non coding sequence (be it non coding DNA, or a duplicated, inactivated gene). We have no restrictions from negative selection. We can go anywhere.

    The starting sequence has no special property: we just know, a posteriori, that it is not homologous to the sequence to be found, because as we have said there is no detectable homology in the pre-vertebrate proteome.

    Of course, some random non coding sequence could have some homology to the final sequence, but why? By sheer luck? That it as unlikely as finding the sequence by a random walk!

    So, let’s say that we start from an unrelated, non funtional sequence, and that we go on by a random wlak, where each step can go not only in any direction, but also as far from the starting sequence as we like.

    This is really the best scenario for RV, IMO.

    OK, my simple point is: If we have a search space of 2^1049, using the blast bit value, which is the more conservative option, or even bigger (the brute search space for a specific 525 AAs sequence is 2269 bits), and we start a random walk from any unrelated sequence, then each sequence in the search space has approximately the same probabilities to be found.

    So, the probabilities I use in my reasonings are the best probabilities for free RV in a darwinian scenario.

    Any improvement to such a devastating truth can come only from:

    a) Adding intelligent specific information to the search (for example by intelligent selection, or by some intelligent control of the parameters of the search).

    b) Natural selection. Which brings us to all the discussion in this thread, and in particular to my challenge at #103.

  165. 165
    Origenes says:

    GPuccio: Excuse me, this new sequence is new. By definition, it has no homology with what existed before.
    So, no sequence with any “nearness” to this specific functional island existed before vertebrates.

    So, in the specific case of SATB1, and the other protein sequences you have written about, recombination can be ruled out as part of any explanation. Good to know. Very important to point this out.

    Would you care to comment on the general objection against recombination: ‘it can’t pieced together like LEGO bricks’? It seems very important to me.

    Axe: “For those elements to work as robust modules, their structure would have to be effectively context-independent, allowing them to be combined in any number of ways to form new folds.”
    [see also #155]

    One more related thing, I have a concern regarding computational biology: from the paper, referenced by Mung #156 :

    In Compositional Evolution, Richard Watson uses the tools of computer science and computational biology to show that certain mechanisms of genetic variation (such as sex, gene transfer, and symbiosis) allowing the combination of preadapted genetic material enable an evolutionary process, compositional evolution, that is algorithmically distinct from the Darwinian gradualist framework.

    Now I can imagine that things combine much more easily in some computer program than in reality where things do not act as LEGO bricks. Do these guys acknowledge this? I felt the same concern when reading Andreas Wagner — another ‘computational biologist’.

  166. 166
    Eugene S says:

    GPuccio

    “So, we have no problems of “near neighbourhood search” or “large neighborhouud search” here. We just have a pure random walk.”

    If that is true, then there is no issue. Correct me if I am wrong, the walk is really done not in proteome space but in genome space, right? Maybe it is a really daft question, but can there be situations where what is near in terms of the genome, is distant in terms of proteome?

    I am not trying to question the ID hypothesis, which has a lot of theoretical and practical merit, or your method. It is really impressive stuff to show new complex functionality appear and get conserved across phyla. If I have questions, they are not meant to undermine anything. I want to get things right in my head 😉

  167. 167
    Corey Delvine says:

    Well for starters let me point you to a couple websites that anyone can use to find accurate estimates of lineage branches:
    One zoom tree of life
    and
    Interactive tree of life evogeneao

    While these are obviously not published research papers, I’d bet they are quite accurate based on current knowledge and their ease of use means they can help you out with your attempts at science.
    Now, both these sites put the human-spider split at 560-590mya, while the human-shark split was 460mya. So the time difference appears to be approximately 100 million years, not 30.

    You can convert bits into a probability all you want, but “evolution has to sample every combination or just about every combination of 2^1049” (what you are basically saying) is a terrible assumption.
    That may end up being another discussion entirely, but for now there is no reason we can’t leave bits as bits and convert it into a rate based on the time between branching events.

    Now, my question, that apparently you couldn’t understand, was what does this rate look like for other proteins, in other species, and with different time gaps.
    Is 35 bits/million years close to the average for other proteins?
    What you are doing is basically subsampling all protein evolutionary paths and reporting what seems like extreme cases in bit-jumps.
    What I am asking is are these bit-jumps significantly different from other protein evolution events, or are you just pointing to large gaps in time where we don’t have sequences from because of extinction events and saying “look at this huge number!”

  168. 168
    ET says:

    Corey- No one knows what makes an organism what it is. Evolutionists have all of their hopes on organisms being the sum of their genomes but science has yet to demonstrate such a thing.

    Then there is the problem of all of the anatomical and physiological differences observed between two allegedly related species, like humans and chimps. No one can link those differences to the genetics.

    And AGAIN, ID is not anti-evolution. Blind watchmaker evolution only has sheer dumb luck and science doesn’t do that

  169. 169
    Dionisio says:

    I like the discussion between EugeneS and gpuccio.

    They’re refreshing some interesting concepts that should be well understood and very rigorously presented.

    Origenes, ET, DATCG, Mung, cmow bave contributed with interesting comments too.

  170. 170
    Eugene S says:

    GPuccio

    Re your comments 5 & 6. I have finally got to reading them. Great stuff, much appreciated. They provide necessary detail in response to my questions. Incidentally, I was also interested in the role of genetic drift in terms of probabilistic resources. You touched on it there as well.

    BTW, yes, I understand the business about the already existing function in that paper! They did not find any new function. Instead they simplified the existing function. Evolutionary motion in practice is isolated by islands of functionality.

    My questions regarding large neighbourhoods have now become less of a concern: too many intelligent things need to happen in order for them to be of any biological value.

    Everything clicks back in place. Thanks.

  171. 171
    Dionisio says:

    @169 error correction,

    “[…] have contributed with interesting comments too”

  172. 172
    Dionisio says:

    Eugene S @170:

    […] too many intelligent things need to happen in order for them to be of any biological value.

    Exactly.

    However, we should note that what gpuccio and you have discussed is about getting the complex functional specified information in the proteome. That’s just like the cacophony we hear when the orchestra musicians are tuning their instruments. The actual ballet starts after that.

  173. 173
    gpuccio says:

    Corey Delvine:

    Well for starters let me point you to a couple websites that anyone can use to find accurate estimates of lineage branches:
    One zoom tree of life
    and
    Interactive tree of life evogeneao

    While these are obviously not published research papers, I’d bet they are quite accurate based on current knowledge and their ease of use means they can help you out with your attempts at science.

    Thank you. Of course, I have looked at similar sites, maybe not exactly the same.

    Then you say:

    Now, both these sites put the human-spider split at 560-590mya, while the human-shark split was 460mya. So the time difference appears to be approximately 100 million years, not 30.

    Haven’t you read my post #161, in answer to you? Or haven’t you understood it? Or are you simply ignoring what I say?

    I have never considered the spider – humans split, which would correspond to the split between protostomes and deuterostomes.

    In my post, I say, rather explicitly and, I believed, clearly:

    The second time, 440 million years, is, similarly, the time that pre-vertebrates diverged from the human lineage. That corresponds to the time of the split between non vertebrate chordates (IOWs, Cephalochordata and Tunicata) diverged from vertebrate chordates (IOWs, the common ancestor of cartilaginous fish and bony fish).

    The relevant splits here are:

    Protostomes-deuterostomes

    Other Deuterostomes (echinoderns, emichordates)-Chordates

    Non vertebrate chordates (Cefalochordata, Tunicata) – Vertebrates

    The last split is the one we are interested in.

    Please, read that post (again).

    Moreover, you should understand that the timing of splits can be assessed in at least two major ways:

    a) Fossils

    b) Molecular reasonings, based on various kinds of molecular clock analyses, often rather discordant.

    I have tried to stick to fossils as much as possible.

    However, I am not really interested in discussing the time window. Do you prefer 100 million years, instead of 30 millions? Be my guest. It makes no real difference.

    You see, 1049 bits is too much even for a moderately dimensioned multiverse.

    Dembski’s Universal Probability Bound is 500 bits: that represents a reasonable assessment of the greates search space that can be traverse by a random search using the whole universe as a computer, and quantum basic units of time and space as computationsal units.

    1049 bits is hugely greater than that!

    So, do you prefer 100 million years? No problem. Do you prefer 1 billion years? No problem. No problem at all. Do as you like.

    Then you say:

    You can convert bits into a probability all you want, but “evolution has to sample every combination or just about every combination of 2^1049” (what you are basically saying) is a terrible assumption.
    That may end up being another discussion entirely, but for now there is no reason we can’t leave bits as bits and convert it into a rate based on the time between branching events.

    I am not converting anything. The bit values I am using are a measure of probability. I hope you understand that.

    And I am not “converting” them into a rate. You have introduced, very inappropriately, the concept of a “rate” (whatever you mean with that word).

    What I am saying is extremely simple, and you should be able to understand it:

    A functional sequence with a complexity of 1049 bits appears in a time window of 30 million years (or 100, or whatever you like). The probability of getting that result by a random walk is practically zero. Therefore, the hypothesis is not an acceptable scientific explanation, and must be firmly rejected.

    Is that clear? As you can see, I am not converting anything, and I am not using any “rate” at all.

    By the way, in your post #167 I can see no mention of the fact that you had intepreted an exponential measure as a linear measure, and made wrong computations with it. Why are you keeping silent about that?

    A simple statement of any kind, like:

    a) I apologize, that was a mistake

    or

    b) No, you are wrong. That is a linear measure, and it is not exponential!

    or

    c) I was just joking!

    would be appreciated. But mere silence?

    Then you say:

    Now, my question, that apparently you couldn’t understand, was what does this rate look like for other proteins, in other species, and with different time gaps.
    Is 35 bits/million years close to the average for other proteins?

    I still don’t understand. What is this “rate” you are speaking of?

    You insist with the figure of 35 bits/million years, which, as I have explained, is only the result of your gross error, the division of an exponent by a linear time measure.

    I cannot imagine what the “average” of a wrong and meaningless number could be for other proteins.

    Then you say:

    What you are doing is basically subsampling all protein evolutionary paths and reporting what seems like extreme cases in bit-jumps.

    Yes. I have a database with those values for all proteins in the human genome.

    Then you say:

    What I am asking is are these bit-jumps significantly different from other protein evolution events, or are you just pointing to large gaps in time where we don’t have sequences from because of extinction events and saying “look at this huge number!”

    In the following OP:

    https://uncommondescent.com/intelligent-design/the-amazing-level-of-engineering-in-the-transition-to-the-vertebrate-proteome-a-global-analysis/

    I have given some general results of my methodology as applied to the whole human genome. Please, read it.

    (By the way, I have just realized that for some strange reason the Figures in that OP had been lost! I have fixed that.)

    Now, look at Table 1 and Figure 1. There you can find the mean values for all the groups I have considered.

    If you look at Figure 1, you can find the mean values of human homology for all the proteins in the human preoteome, expressed as bits per aminoacid site (IOWs, corrected for protein length).

    As you can see, the biggest jump at all is between pre-vertebrates and vertebrates (non vertebrate deuterostomia – cartilaginous fish).

    For comparison, there is only a really trivial mean jump between reptiles and and mammals (crocodiles – marsupials).

    Of course, these are mean values. For individual proteins, there is of course great variance.

    And yes, the examples I have given are certainly “extreme” cases, but there are a lot of them. However, the mean jump in vertebrates is of about 189 bits per protein

    The total figure is 1.7+ million bits.

    Of course, there are a lot of proteins which have similar homology in pre-vertebrates and first vertebrates. It is obvious that not all existing proteins have been highly engineered in the transition to vertebrates: only a very big subset of them. A future OP will discuss what are the specific sets of proteins that are engineered in that transitions.

    For your convenience, here is a more detailed distribution of the information jump in absolute bits in that transition:

    Mean: 189.5817

    SD: 355.6362

    IQR: 224

    Median: 99

    90th percentile: 486

    The distribution has been evaluated on 19564 protein sequences.

    As you can see from the 90th percentile value, 10% of all human proteins have an information jump of more than 486 bits in the transition to vertebrates!

    Are you still thinking that these are only “extreme cases”?

    Remember, Dembski’s UPB (for our whole universe) is 500 bits.

    I hope this answers you questions.

    And, by the way, thank you for humbly admitting that I am making “attempts at science”. Coming from you, it’s a big compliment, which I will add to the other precious acknowledgements I have already received from you.

    You are really spoiling me! 🙂

  174. 174
    gpuccio says:

    Eugene S:

    Thank you for your comments! 🙂

    I am happy that we agree on the big picture.

    I will be very happy to comment on the points you raise as soon as possible (now I really don’t have the time).

    Thank you for everything. 🙂

  175. 175
    Mung says:

    I am not converting anything. The bit values I am using are a measure of probability.

    The Shannon Measure of Information really is that simple.

    Eight choices with equal probability? 3 bits. 2^3 = 8.

  176. 176
    Origenes says:

    //follow-up #165//

    In #165 I asked the question: “Now I can imagine that things combine much more easily in some computer program than in reality where things do not act as LEGO bricks. Do these guys acknowledge this?”

    Rereading Andreas Wagner confirms my concern:

    The evolutionary algorithms that mimic biological evolution are powerful tools, but they’re still missing something. They are still deficient in the recombination department so central to biological innovations.31 Nature is better at recombination, much better, for one simple reason: standards.

    The absence of such standards makes recombination more difficult in technology, which often uses ingenuity as a substitute. It took plenty of ingenuity to recombine a compressor, a combustion chamber, and a turbine to lift a plane’s many tons of metal into the air. …

    … The first bicycles combined three, the wheel, the lever, and the pulley. All of these combinatorial innovations required ingenuity.
    … most technologies are deficient in a certain kind of standard, the one that allows nature to combine the old to make the new. Nature needs these standards, precisely because it does not have the ingenuity of human inventors.

    Even though each amino acid has a different shape, they can all connect in the same way, because they share a universal interface. And this standard, used by all organisms, has made life as we know it possible. It allows nature to cobble together—blindly, without any ingenuity—the astronomical numbers of genotypes needed to find innovations.
    Standards that make recombination mindlessly easy do not just exist in proteins. RNA strings also use a standardized chemical bond to link their parts. Furthermore, life’s standard of information storage—DNA—allows bacteria to exchange genes and create new metabolisms from new combinations of the enzymes they encode. And finally, regulation circuits use a standardized way to regulate genes, based on the principle that regulator proteins can bind specific short words on DNA, allowing nature to combine old regulators into countless new circuits by changing these words.32 If we could take a small number of different objects, create a standard way to link them, and recombine them into every conceivable configuration—mindlessly—our powers to innovate could be just as immeasurable as those of nature.
    Such standardization is clearly not beyond human technology: Our hardworking Lego blocks hint at it, and so does a much older human technology.
    [Andreas Wagner, ‘The Arrival of the Fittest’]
    [emphasis added]

    What is completely lacking in this story is the three-dimensional structure of proteins. It might be “mindlessly easy” to combine DNA and RNA sequences, but lets not forget that they do represent non-standard three-dimensional structures.
    Again Axe:

    “For those elements to work as robust modules,” explains Axe, “their structure would have to be effectively context-independent, allowing them to be combined in any number of ways to form new folds.” In the case of protein secondary structure, however, this requirement is not met.

    And again this:

    A further issue relates to interface compatibility. The domain shuffling hypothesis in many cases requires the formation of new binding interfaces. Since amino acids that comprise polypeptide chains are distinguished from one another by the specificity of their side-chains, however, the binding interfaces that allow units of secondary structure (i.e. ?-helices and ?-strands) to come together to form elements of tertiary structure is dependent upon the specific sequence of amino acids. That is to say, it is non-generic [non-standard! — Origenes] in the sense that it is strictly dependent upon the particulars of the components. Domains that must bind and interact with one another can’t simply be pieced together like LEGO bricks.
    [Jonathan M.]

    It may be telling that the term “binding interface” is nowhere to be found in the book written by computer biologist Andreas Wagner.

  177. 177
    gpuccio says:

    Eugene S and Origenes:

    The subject of recombination is vast and complex.

    I will try to address some of the points you have made in this interesting discussion, but I would like to start with a general premise.

    Recombination is a concept that includes many different mechanisms and scenarios. So, it is important to distinguish between them.

    In general,I would say that we can consider recombination any situation where already existing information, usually functional, is “moved” to a different context, while the information sequence itself remains mainly unchanged.

    Of course, the idea is that the change in context generates some relevant change in function, possibly a gain in function or the appearance of some new function.

    Now, even in design scenarios it is certainly possible to arrange existing modules in new, original ways, and derive new function from that process. Indeed, object-oriented programming is exactly about that: being able to use the same modules in different contexts, usually changing only the interface.

    Of course, in a design scenario the rearrangement is guided by design principles, and is therefore in itself an addition of functional information: let’s say that, in the search space of all possible rearrangements, only a very tiny minority will be functional.

    But let’s see how different are the scnarios of genetic recombination:

    a) Sexual recombination (meiotic recombination). This usually realizes an exchange of homologous alleles between non identical chromosomes (one paternal, one maternal). This is perhaps the best example where variants of the same basci information (for example different alleles of a proteins) are remixed in the progeny.

    Two important facts about this are:

    1) The process is not an error process, but rather a highly adaptive process, and is linked to the nexhanisms of DNA repair:

    From Wikipedia:

    Genetic recombination is catalyzed by many different enzymes. Recombinases are key enzymes that catalyse the strand transfer step during recombination. RecA, the chief recombinase found in Escherichia coli, is responsible for the repair of DNA double strand breaks (DSBs). In yeast and other eukaryotic organisms there are two recombinases required for repairing DSBs. The RAD51 protein is required for mitotic and meiotic recombination, whereas the DNA repair protein, DMC1, is specific to meiotic recombination. In the archaea, the ortholog of the bacterial RecA protein is RadA.

    2) The remix of allele seems to be an adaptive mechanism that contributes to generate diversity in a population. This is really trivial: we know very well how children are in many ways a “shuffling” of traits present in parents. While this is certainly due to the fact that they inherit one copy of a chromosome from the father and one from the mother (which, in itself, is a form of recombination), meiotic recombination contributes to change the linkage between trais, and therefore the final diversity in the population.

    Races (in dogs, cats, and so on) are probably a good example of diversity by shuffling of alleles, although other mechanisms are probably imnportant too.

    The important point is that here the functional unit, the protein, remains the same, but different variants of the same protein (different alleles) are shuffled around.

    b) Exon shuffling. This is different, because here parts of existing proteins are shuffled, in new combinations. This is a very complex subject, and I will discuss it in more detail later.

    I would like to anticipate, however, the important role payed by transposons in this process. Like meiotic recombination, this seems to be a very active process, and not only the result of errors.

    c) Whole parts of existing information can also be transferred due to errors in mitosis; translocations and inversions are a goos example. These mechanisms are very important in tumors, for example, or in congenital malformations.

    d) Of course, duplications and deletions (indels) are forms of recombination too. Whole chromosomes can be dupliated or lost, or parts of them.

    e) HGT is a form of recombination too: an active process, again.

    f) A very interesting form of functional recombination to generate diversity from a limited genetic repertoire is V(D)J recombination in somatic immune cells, a process that I have debated here as a wonderful form of biological intelligent engineering in algorithmic form:

    https://uncommondescent.com/intelligent-design/antibody-affinity-maturation-as-an-engineering-process-and-other-things/

    So, this is just to show how complex the subject is.

    I will try to address some of your specific points in later posts.

  178. 178
    gpuccio says:

    Eugene S:

    I would like to start with this question of yours, which is simple enough:

    Correct me if I am wrong, the walk is really done not in proteome space but in genome space, right? Maybe it is a really daft question, but can there be situations where what is near in terms of the genome, is distant in terms of proteome?

    Of course, the walk is in the genome space, because it’s there that the variation takes place.

    But variations are translated into variations of the proteome, by the genetic code.

    So, there is a very strict parallelism between the two things.

    It is important ot remember, however, that the genetic code is redundant, and that has important consequences, which in general do not change very much the correspondence between genomic space and protein space. These are:

    a) Synonymous mutations change the genomic sequence, but not the proteomic sequence. The redundancy, however, is mainly linked to the third letter of the codon. Synonymous mutations are important because they are the foundation of the Ka/Ks ration to evaluate the effects of NS. However, while they certainly can have some functional relevance, in general we can accept that most of the functional variatoon is due to non synonymous mutatios, or in general to any variation which affects the protein sequence.

    b) Different variations at genomic level can give the same variation at protein level, due to the redundancy.

    That said, the fact remains that function, and therefore selection, depend on the protein sequence, and so, even if variation starts an the genomic level, it must be evaluated mainly for its effect on the protein sequence. At least for protein coding genes.

    c) There are some situations where, as you say: “what is near in terms of the genome, is distant in terms of proteome”. For example, frameshift mutations. A single insertion or deletion in the genomic sequence can change completely the translation of the sequence. Also, the acquisition or the loss of a start codon, or of a stop codon, can determine the loss of a protein, ot its truncation, or the sudden transcription and translation of a non coding sequence.

    As you can see, all those situations are the equivalent of a sudden jump into the vastity of the ocean of randommprotein sequences by a simple genetic modification: any esisting function, in these cases, will be lost.

    However, the opposite is not true: what is near at the level of the proteome is also near at the level of the genomic sequence. The only exception being synonymous differences.

    IOW, if two proteins have the same AA sequence, they must also have the same genomic sequence, except for the redundancy given by synonymous mutations.

  179. 179
    Dionisio says:

    gpuccio @177:

    Did you dare to associate object-oriented programming with biological systems?

    Perhaps it’s just another illusion, like design and consciousness. 🙂

  180. 180
    gpuccio says:

    Dionisio:

    I like bold metaphors.

    If they can use cars and planes, why can’t I rely on OOP? 🙂

  181. 181
    gpuccio says:

    Origenes at #176 and Eugene S:

    Well, form the premise that I have made in my previous posts, it seems that the most promising mechanism to generate new functions seems to be exon shuffling.

    Before dealing with it in some depth, however, I would like to comment on the points raised by Origenes at #176.

    I will be very sincere now: I have never liked Andreas Wagner.

    Not that I have read much of him (and I see no reason to do that), but the little I have read, or that has been quoted to me, is well beyond tolerability. Indeed, his words are usually a concentrate of nonsense.

    The passage quoted in #176 is no exception. Really, I would say it is the perfect incarnation of Wagner’s style and way of “reasoning”.

    While I was reading it, I could not believe in what I was seeing.

    So, nature has no ingenuity, but it has standards!

    And what are the standards?

    If I understand well, it’s the fact that nucleotides can combine in a sequence, and aminoacids can combine in a sequence, by a specific biochemical binding that is the same for all nucleotides (and for all aminoacids).

    So, he has discovered the chain-joins of sugar and phosphate in nucleotide sequences, and the peptide bonds in protein sequences, and realized for the first time how that makes biological polymers similar to Lego blocks!

    This is utter nonsense.

    It would have been better to say that those chemical bonds make biological polymers similar to words, where nucleotides and aminoacid are letters.

    Have you ever noticed how wrods are similar to Lego blocks, and how easy it is to build poems and novels without any ingenuity, given the simple facts that letters can be easily joined in a sequence?

    The only “mindlessly easy” thing here is the way Wagner recombines simple concepts to get meaningless ideas.

    That said, I must confess that I am not able to get any acceptable and functional result from Lego blocks, even with all my ingenuity! That is certainly my problem in understanding Wagner! 🙂

    Instead, I absolutely agree with what Axe says. It is perfectly true. In particualr, it is true that secondary structures in themselves, for example alpha helices, while certainly representing some level of organization, could be considered as “functional modules”. They are not.

    The relationship between secondary structure and tertiary structure is too complex to think that secondary structures are functional modules that can be arranged as Lego bricks. That is not true.

    Tertiary structure is so complex that we cannot really anticipate it with efficiency or reliability by our computational resources and our human ingenuity. Try that with Wagner’s mindless easiness!

    From Wikipedia:

    Tertiary structure

    The practical role of protein structure prediction is now more important than ever. Massive amounts of protein sequence data are produced by modern large-scale DNA sequencing efforts such as the Human Genome Project. Despite community-wide efforts in structural genomics, the output of experimentally determined protein structures—typically by time-consuming and relatively expensive X-ray crystallography or NMR spectroscopy—is lagging far behind the output of protein sequences.

    The protein structure prediction remains an extremely difficult and unresolved undertaking. The two main problems are calculation of protein free energy and finding the global minimum of this energy. A protein structure prediction method must explore the space of possible protein structures which is astronomically large. These problems can be partially bypassed in “comparative” or homology modeling and fold recognition methods, in which the search space is pruned by the assumption that the protein in question adopts a structure that is close to the experimentally determined structure of another homologous protein. On the other hand, the de novo or ab initio protein structure prediction methods must explicitly resolve these problems. The progress and challenges in protein structure prediction has been reviewed in Zhang 2008.[26]

    Emphasis mine. IOWs, the prediction of tertiary structure from sequence is so difficult, that the best way to make it simpler is to rely on similar proteins whose structure is already known by other methods (like crystallography).

    So, Axe is absolutely right, and Wagner is absolutely wrong.

  182. 182
    gpuccio says:

    To all:

    Nobody has tried to answer my challenge (two very simple questions) at # 103 (almost 80 comments ago).

    And Corey Delvine is still silent about the “problem” of the exponential nature of information measures.

    Ah, what a calm life!

    For the convenience of all, I paste here again the challenge:

    Now, the challenge.

    Will anyone on the other side answer the following two simple questions?

    1) Is there any conceptual reason why we should believe that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    2) Is there any evidence from facts that supports the hypothesis that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

  183. 183
    Origenes says:

    Challenging all Darwinians:

    Will anyone on the other side answer the following two simple questions?

    1) Is there any conceptual reason why we should believe that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    2) Is there any evidence from facts that supports the hypothesis that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    Gordon is obviously invited. Corey is invited. Anyone on the other side is invited.

    [GPuccio]

    – – – – –
    GPuccio, thank you for writing on recombination.

    GPuccio: Tertiary structure is so complex that we cannot really anticipate it with efficiency or reliability by our computational resources and our human ingenuity. Try that with Wagner’s mindless easiness!

    This confirms my suspicion wrt Wagner’s ideas. Again, thanks!

  184. 184
    Dionisio says:

    Congratulations gpuccio!

    Popular Posts (Last 30 Days)

    Violence is Inherent in Atheist Politics (1,794)

    Howling Darwinists (1,466)

    What are the limits of Natural Selection? An interesting… (1,420) [since October 5]

    Selective Horrid Doubt (1,313)

    Wolf-Ekkehard Loennig Falsifies Darwinism (1,213)

  185. 185
    gpuccio says:

    Dionisio and Origenes:

    Yes, I am happy of the discussion here.

    Only, no joy from the challenge.

    Maybe I could give some help and encouragement. Such as suggesting possible answers:

    a) Yes, there are conceptual reasons and/or evidences from facts that support that idea. Here they are…

    b) No, there are no conceptual reasons and/or evidences from facts that support that idea, but why should we need them? Of course, neo-darwinism is right just the same.

    c) That is not a scientific question, because it makes no predictions.

    d) Those questions violate methodological naturalism: the concept itself of deconstruction is not natural at all!

    e) Ladders are not an appropriate metaphor for reproducting systems.

    And, finally, my favourite one:

    f) You just don’t understand how evolution works! 🙂

  186. 186
    Mung says:

    Dear gpuccio,

    I use one of those sonic toothbrushes. I discovered a sound-sensitive spot on my tooth that was not there previously. I am well on my way to an ear. Stay tuned.

  187. 187
    Dionisio says:

    RodW @152:

    Did you read the comment @157?

    Did you understand it?

    Is that issue clear now?

  188. 188
    Dionisio says:

    Mung @186:

    “I am well on my way to an ear. Stay tuned.”

    You may want to review the comments @160-162 in the following link, which could help with your research:

    https://uncommondescent.com/intelligent-design/the-beautiful-mechanism-by-which-an-egg-becomes-an-embryo/#comment-641297

    Note that the biological process underlying the inner ear formation is fully documented in all details and it seems pretty simple, like a Lego Dupplo toy.

    🙂

    PS. BTW, if you still have questions, you may direct them to gpuccio’s politely dissenting interlocutors within this thread. Note that they all seem to know pretty much everything, or at least can figure it out quickly.
    🙂

  189. 189
    gpuccio says:

    Mung:

    I knew I could count on you! 🙂

  190. 190
    gpuccio says:

    Dionisio:

    Yes:

    “Sculpting the labyrinth: Morphogenesis of the developing inner ear”

    is specially fascinating!

    The advent of powerful non-invasive live-cell imaging and labelling techniques is revolutionising the developmental biology field. In particular, light-sheet microscopy now enables long time-lapse imaging deep in tissues, in all spatial dimensions, with high temporal resolution and little phototoxicity [201–203], providing unprecedented new information on the kinetics, dynamics, and 3D architecture of morphogenesis at the cellular, tissue or organ level. In addition, the physics of morphogenesis is beginning to be elucidated by combining visualisation of fine-grained sub-cellular details with novel non-invasive nano/picoscale technologies for mechanical manipulation of tissues. Some of the tools for monitoring mechanics forces are laser-cutting devices, micropipettes to analyse mechanical and adhesive properties of cells and tissues, and, finally, molecular force sensors. A plethora of relevant data on the impact of forces, tensions, pressure and flows in embryogenesis has been published recently focusing on gastrulation, heart and endothelial development, among many others

    Emphasis mine.

    Revolutionazing, not just “changing”.

    A plethora, not just “a lot”.

    Wow!

    Shall I say it? OK:

    “Complex functionally specified informational complexity”! 🙂

  191. 191
    Dionisio says:

    gpuccio @190:

    Agree.

    The text you quoted is written in exciting terms:

    “…revolutionising the developmental biology field”

    “…unprecedented new information on the kinetics, dynamics, and 3D architecture of morphogenesis at the cellular, tissue or organ level.”

    “…the physics of morphogenesis is beginning to be elucidated…”

    “A plethora of relevant data on the impact of forces, tensions, pressure and flows in embryogenesis…”

    On the second half of the second decade of the 21st century?

    The most fascinating discoveries are still ahead.

    BTW, the ‘complex complexity’ phrase at the end of the comments is an understatement. 🙂

    Emphasis added.

  192. 192
    gpuccio says:

    Dionisio:

    Nobody has taken the challenge.

    It seems that nobody is interested in the problem of deconstructing complex functional information into naturally selectable steps.

    And yet, neo-darwinists are always ready to defend the role of NS in building complex functional information by some imagined pathway, never specified.

    Do they even understand that the existence of such a pathway critically depends on the problem of deconstruction?

    If they are so certain that those pathways exist, they should be equally certain that complex information can be desconstructed into a ladder of simple naturally selectable steps.

    That should be a pillar of their convictions, of their cognitive assurance.

    So, why don’t they try to answer my two simple questions?

  193. 193
    Origenes says:

    The Darwinians here are all cowards.

    GPuccio: why don’t they try to answer my two simple questions?

    I say this with great reluctance, because I make it a goal to see the best in everyone, but it is the only explanation I can think of.

  194. 194
    Dionisio says:

    Maybe the reason nobody has taken gpuccio’s challenge is simply that he doesn’t ask honest questions, whatever that means?

    🙂

  195. 195
    Dionisio says:

    Maybe the reason nobody has taken gpuccio’s challenge is simply that he doesn’t understand evolution, whatever that means?

    🙂

  196. 196

    So, why don’t they try to answer my two simple questions?

    There seems to be a lot of that going around lately.

  197. 197
    Dionisio says:

    UB, that’s a valid observation. Agree.

  198. 198
    Dionisio says:

    Popular Posts (Last 30 Days)

    Violence is Inherent in Atheist Politics (1,935)

    What are the limits of Natural Selection? An interesting… (1,520) [since October 5]

    Howling Darwinists (1,468)

    Selective Horrid Doubt (1,315)

    Sweeping the Origin of Life Under the Rug (1,000)

    The latest stats on visitors may indicate that lots of folks out there are anxiously expecting that some politely dissenting interlocutor answers gpuccio’s easy questions.
    But I think it ain’t gonna happen anytime soon…

    🙂

  199. 199
    Corey Delvine says:

    Ok, I guess I better understand how your game works, Gpuccio.

    I’ll try to make things simple:

    Do you really think that protein evolution has to explore all, or even close to all of the amino acid sequence space you are claiming?

  200. 200
    ET says:

    Corey:

    Do you really think that protein evolution has to explore all, or even close to all of the amino acid sequence space you are claiming?

    There isn’t any exploration. It is all just sheer dumb luck. How much luck can science allow?

  201. 201
    Origenes says:

    Corey @199

    Not if complex protein functions can be deconstructed into simpler, naturally selectable steps — see GPuccio’s challenge @103.
    But if such a ladder does not exist, then protein evolution is, at best, a blind search in a monstrous sequence space.

  202. 202
    Dionisio says:

    To Whom This May Concern:

    EugeneS @170:

    […] too many intelligent things need to happen in order for them to be of any biological value.

    @172:

    […] we should note that what gpuccio and you have discussed is about getting the complex functional specified information in the proteome. That’s just like the cacophony we hear when the orchestra musicians are tuning their instruments. The actual ballet starts after that.

    https://uncommondescent.com/intelligent-design/what-are-the-limits-of-natural-selection-an-interesting-open-discussion-with-gordon-davisson/#comment-641223

  203. 203
    Corey Delvine says:

    Evidence for a “ladder of protein evolution” is all over the literature, whether it’s a basic biology book or a research journal.

    Maybe you guys should try actually opening one of these up.

  204. 204
    ET says:

    Corey- You don’t have a mechanism capable of producing any proteins. Your position has to start with proteins already in existence. Then to get a new protein from an existing one takes multiple specific changes and with your proposed mechanism there just isn’t enough time available.

    Maybe you should try opening your eyes.

  205. 205
    Dionisio says:

    Folks, please, let me know when you’re done trying to figure out how to recruit and train the required dancers (aka functional protein domains), so we can discuss the actual ballet choreography.
    🙂

    We have to rehearse for the début presentation early next year of the new ballet titled “biological humpty dumpty”, which combines Latin America and Africa folklore dance + Italian opera + Broadway show + classic symphonies + rock music + country music + disco music + Hollywood cowboys movies + the whole nine yards. 🙂

  206. 206
    Dionisio says:

    @205 addendum

    This is kind of old (at least 12 years?), but perhaps still somehow related to our “biological humpty dumpty” ballet project:

    http://www.cell.com/current-bi.....0278-2.pdf

    🙂

  207. 207
    Dionisio says:

    gpuccio,

    Please, accept my apologies for posting off-topic comments @205-206.

    I got confused by the hogwash written @203, but that’s an unacceptable excuse.

    🙂

  208. 208
    Origenes says:

    Corey Delvin: Evidence for a “ladder of protein evolution” is all over the literature, whether it’s a basic biology book or a research journal.

    If that is the case, then perhaps you can provide us with an example. One single example suffices to prove your point.

  209. 209
    gpuccio says:

    Corey Delvin:

    OK, you have not taken my challenge (yet!), but you are the only one on the other field speaking about it, at least! You are a treasure! 🙂

    So, you certainly deserve in depth answers to the points you raise.

    Let’s start with your #199.

    Ok, I guess I better understand how your game works, Gpuccio.

    That’s good. Understanding is always a precious thing. I don’t know what you mean with “your game”, but I take it as a compliment: I do like games! 🙂

    I’ll try to make things simple:

    That’s even better: simplicity is a very precious thing.

    Do you really think that protein evolution has to explore all, or even close to all of the amino acid sequence space you are claiming?

    This is a good question, and it deserve a detailed answer. I apologize in advance if it is not simple enough for your expectations. 🙂

    OK, the answer is: a random walk needs not explore all, or close to all the search space. But it certainly needs to explore enough of it to make finding the target space reasonably likely.

    Let’ go more in depth.

    Let’s say that we have a functional target space whose functional information (the target space/search space ratio) is 500 bits. Dembski’s UPB.

    That means that a random walk, each time it explores a new state, has a probability of 1:2^500 of finding the target space.

    OK?

    Now, we must consider the second inportant element: the probabilistic resources of the system. IOWs, how many new states can be tested by the system in the given time?

    I will take a very generous threshold for that: 140 bits.

    That is an approximation derived from the following parameters, which refer to the most favorable situation for a random walk: a huge population (bacteria) with fast reproduction rates and 5 billion years of evolutionary time available (more than earth’s lifetime). As follows:

    Total number of bacteria on earth: 5E+30

    Mean mutation rate per genome per cell generation: 0.003

    Mean number of generations per day: 30

    Days in 5 billion years: 1.825E+12

    Multiplying the 4 values, we get:

    8.2125E+41

    that is:

    2^139.2368734

    IOWs, 140 bits.

    OK?

    That meas that a random walk, utilizing a very generous estimation of all the probabilsitic resources of our planet in the biological world, can test, at most:

    2^140 different states of a bacterial genome.

    So that will be our upper limit of biological probabilistic resources on our planet, in the following discussion.

    So, now we have that, to find a specific functional island of a specific protein which has 500 bits of functional information, using all the probabilistic resources of our planet we can explore, at most:

    2^140 states

    in a search space of:

    2^500

    That is a fraction of the search space equivalent to:

    1:2^360

    IOWs, the probability of finding that target space, using all the probabilistic resources of the planet in its whole existence, will be about:

    2^-360, which is more or less equal to:

    1.0e-108

    Now, that is our p-value for the null hypothesis that our protein is the result of a random walk.

    Is that improbability small enough to reject that hypothesis?

    You bet! 🙂

    Of course, we will not even take in consideration the conventional threshold of 0.05 to reject a null hypothesis. That would really be ridiculous.

    So, let’s go to the standards of physics, which are much more stringent.

    In physics, the requisite for a big, important discovery is 5 sigma. That was the standard for the Higgs boson, for example.

    What does that mean? It means that we reject the null hypothesis if the probability of our data is lower than that of the upper tail above 5 standard deviations in the normal distribution.

    The corresponding p-value for 5 sigma is:

    2.866516e-07

    Our p-value is (about):

    1.0e-108

    IOWs, our p-value is about 10^101 times lower than the very strict threshold used for fundamental discoveries in physics!

    So, can we reject the null hypothesis of a random walk as an explanation for our protein?

    As I have already said, you bet! 🙂

  210. 210
    gpuccio says:

    Corey Delvin:

    In your post #203 you say;

    Evidence for a “ladder of protein evolution” is all over the literature, whether it’s a basic biology book or a research journal.

    OK, you are still not taking my challenge.

    However, while you don’t touch at all my question number 1, you make some vague comment about my question number 2.

    If I understand well the vast wisdom in your long post, you are saying that it is very easy to answer my second question, because:

    “evidence for a “ladder of protein evolution” is all over the literature”.

    Well, the most obvious answer has already been given by Origenes (thank you, again! 🙂 ):

    “If that is the case, then perhaps you can provide us with an example. One single example suffices to prove your point.”

    That’s true. If it’s so easy, why don’t you provide any example?

    That said, I will add that I find your expression “a ladder of protein evolution” extremely vague and ambiguous. So, to prevent misunderstandings, let’s look again at the second question in my challenge, as clearly stated in my post #103 (more than 100 comments ago!):

    2) Is there any evidence from facts that supports the hypothesis that complex protein functions can be deconstructed into simpler, naturally selectable steps? That such a ladder exists, in general, or even in specific cases?

    As you can easily see, there is no mention here of a “ladder of protein evolution” (your words, whatever they mean).

    The ladder I am speaking of is very well defined: a ladder of simple naturally selectable steps that can build a complex function in a protein.

    But, always to prevent misunderstanding, in the unlikely case that you want to try to answer my challenge, I will try to explain better what I mean, using a hypothetic example.

    Let’s say that we have a protein that appears in natural history at some time, and exhibits 500 bits of new functional complexity, or more.

    As I have shown in my post #173 (to you), that is not at all a rare case. Just considering a single event in natural history, the transition to vertebrates, about 10% of all proteins in the human genome exhibit an information jump like that, or higher. That means about 2000 proteins. You can freely choose among them.

    Now, in our scenario where only RV acts, that is obviously a practically impossible result, as discussed in detail in my post #209.

    Therefore, we have to invoke NS, to lower the probability barriers.

    Now, what is a “simple step”? It is a step which is in the range of what RV can reasonably do, without any help from NS.

    Considering the results of the quoted paper “Waiting for two mutations”, I would advice you to stick to 2 mutations for a simple step. As can be seen from that paper, that is already a very exacting result for RV.

    Or, if you prefer, we can consider a 4 AAs mutation as a “simple step”. That’s not better for you, because the probabilties of a 4 mutations step are exponentially lower than those of a 2 mutations step: that means waiting for much longer times.

    However, you choose.

    So, what you have to show is simple: you have to show that there is evidence for some protein with 500 bits of functional information that a ladder exists such that, starting from any sequence unrelated state:

    a) The information in the final, functional protein can be reached through 250 successive steps, each of them of two mutations (or, if you prefer, 125 steps of 4 mutations each).

    b) Each successive sequence is functional, and is more functional than the previous one, so that the new allele can give a detectable reproductive advantage on the previous allele, and therefore be expanded and fixed by NS.

    Of course, a good answer would be to show not only that there is such evidence for at least one protein, but also that there is evidence that the same can be true in the general case (after all, we have thousands of proteins that we need to explain, not only one! 🙂 )

    But you can certainly start with one.

    This is the challenge. It is not vague. It is very well defined.

    Now you can certainly answer it.

  211. 211
    Origenes says:

    Corey seems to be the kind of darwinian who naively thinks that his theory is well-supported by evidence. One doesn’t see those very often on this forum.
    Meanwhile, his fellow darwinians, the ones who are well acquainted with the fact that evidence for their position is non-existent, remain awfully quiet. One can almost hear the gnashing of teeth.

  212. 212
    Mung says:

    Corey Delvine:

    Do you really think that protein evolution has to explore all, or even close to all of the amino acid sequence space you are claiming?

    Have you considered what the alternatives are?

  213. 213
    Mung says:

    re my 212:

    Miracle #1: By pure happenstance, the search started out in just the right spot in the search space for the search to be successful.

    That would be one possibility.

  214. 214
    ET says:

    Mung @213- It wasn’t even a search. And that makes it even more miraculous.

  215. 215
    kurx78 says:

    gpuccio @210
    “This is the challenge. It is not vague. It is very well defined.”

    According to a famous canadian biochemist and his friends, you are being intellectually dishonest by asking well defined questions 🙂

  216. 216
    gpuccio says:

    kurx78:

    That’s the way with famous canadian biochemists.

    Sometimes they support you (see #74), sometimes they leave you in the gutter! 🙂

  217. 217
    Dionisio says:

    kurx78,

    A couple of years ago I was reprimanded by a distinguished biochemistry professor in this website for asking dishonest questions, whatever that means.
    As gpuccio said, I was left in the gutter, but it was my fault, because I used a tricky word (“exactly”) in a subliminal way instead of writing it in bold characters so that the professor could notice it.
    Shame on me!
    🙂

  218. 218
    Dionisio says:

    gpuccio @209:

    Your explanation is very thorough.

    However, at the end you refer to the standards of physics, but perhaps that’s not fair, because physicists don’t understand Neo-Darwinian evolutionary concepts, whatever they mean?

    🙂

  219. 219
    gpuccio says:

    Dionisio:

    You got it!

    And evolution is a fact, not a theory.

    And even if it were a theory, it’s certainly better than gravitation! 🙂

  220. 220
    Dionisio says:

    gpuccio @210:

    But you can certainly start with one.

    That statement confirms how generous you are to your politely dissenting interlocutors.
    You’ve made it really easy for them to answer your challenge.
    Also that shows that you ask honest questions, hence deserve the support you get from the distinguished biochemistry professor that rightly called my questions ‘dishonest’ a couple of years ago.

    🙂

  221. 221
    Dionisio says:

    gpuccio @219:

    Gravitation theory is very poor in evidence compared to the evidence for a “ladder of protein evolution” that is “all over the literature, whether it’s a basic biology book or a research journal.”

    Physicists would like to have such an abundance of evidence for gravitation, but they can’t have it, because they don’t understand Neo-Darwinian evolutionary concepts. Poor things.

    🙂

  222. 222
    ET says:

    Why is it that only in the field of biology are we to accept that random hits to an existing functioning system does not degrade it? Heck it not only doesn’t degrade it, it made it! And all without evidentiary support? Really?

    And we are the IDiots. 🙄

  223. 223
    Dionisio says:

    ET,

    That’s simply because we don’t understand evolution.

    🙂

  224. 224
    Origenes says:

    ET #222

    Brilliant and concise.

  225. 225
    gpuccio says:

    Origenes:

    “ET #222

    Brilliant and concise.”

    I agree! 🙂

  226. 226
    Corey Delvine says:

    It is clear, gpuccio, that you are very good at pointing to really big and really small numbers.

    You claim to have an idea of search space and target space, but do you really?

    Let’s see.
    In the Szostak paper that’s already been mentioned here, it took just 5 rounds of selection to go from a DNA-binding transcription factor to an ATP-binding protein.

    How on Earth could that be when (according to you and your “research”) with the most generous assumptions, the probability of finding a target space is on the order of 1^-108?

    Who do we beleive?
    Experimental evidence, or gpuccio’s pipe dream number-crunching?

    Now, gpuccio would love for us all to believe that functional islands are so perilously isolated in the land of protein sequence that evolution is something to be scoffed at (along with the decades of research behind it).
    But, let’s take a look at what the (actual) research says:

    Take for instance a recent review on protein folds and their evolution (Clarke et al 2007)

    This paper looked at 1.1 million protein sequences and found that when taking both sequence and structure into account:
    “consitently isolated superfamilies are rare” and “isolated domain superfamilies occur in only 0.16% of protein sequences”

    So, the vast majority of proteins are made up of similar folds?

    Ok, one explanation is that the amount of space evolution actually searches is very small compared to the potential space that it can search and this is due to evolution occurring in a stepwise fashion from existing domains (hence the similarity in all proteins).
    The other explanation is that these are the few and only folds that actually have potential biological function, and as gpuccio wants us to believe, biological function is hard to come by (hence the similarity in all proteins).

    Well which is it?
    The Szostak paper informs us that gpuccio is wrong.
    Starting with a transcription factor that binds DNA, it took only five rounds of mutation and selection to produce proteins that instead bound specifically and significantly to ATP.

    Any reasonable person would re-evaluate their claims in the face of such damning experimental evidence, but not gpuccio!

  227. 227
    Origenes says:

    Corey Delvin @226

    Corey Delvin: The Szostak paper informs us that gpuccio is wrong. … Any reasonable person would re-evaluate their claims in the face of such damning experimental evidence, but not gpuccio!

    You are being ridiculous. In post #62 GPuccio has pointed out that natural selection does not play a role at all in the Szostak paper. It is in fact artificial selection that does all the heavy lifting. Surely, you have to address his objection, before you can claim that the paper shows GPuccio wrong:

    For your convenience here is GPuccio’s objection:

    I have said many times why that paper is methodologically inappropriate as a simulation of NS.

    The main reason, of course, is that it uses AS as a simulation of NS. But it is not AS for an equivalent of what would be naturally selected, like in the rugged landscape paper (where infectivity is selected).

    It is, instead, AS for ATP binding.

    It is important to observe that the original sequence in the original random pool presented only very weak ATP binding, and certainly was not naturally selectable.

    But, even more strikingly, the final protein with strong ATP binding still was not naturally selectable at all.
    [GPuccio]

  228. 228
    Corey Delvine says:

    What exactly is his objection?

    That Szostak didn’t build an entire ecosystem around the protein and observe over millions of years?

    Or is he arguing that there’s no potential for function in ATP-binding?

  229. 229
    gpuccio says:

    Corey Delvin:

    You must be really desperate to resort to the Szostak paper, after all that I have said about it.

    But just a curiosity: do you understand that paper at all?

    You say:

    “Starting with a transcription factor that binds DNA”

    What do you mean? He started “from a library of 6×10^12 proteins each containing 80 contiguous random amino acids”, as clearly stated in the abstract.

    What’s this talk of transcription factors?

    Do you know what you are saying?

    Szostak just selected, by artificial selection realized by ATP-affinity columns, sequences with a weak ATP binding from a random protein library, and artificially mutated them and again selected those with growing affinity for ATP, always by ATP-affinity columns. Obtaining, in the end, a completely useless protein with strong affinity for ATP.

    A very trivial and useless process of intelligent protein engineering, with a bottom up strategy.

    Nothing to do with natural selection, which has absolutely no relation to that process.

    Nothing to do with protein sequence space.

    Wrong methodology and wrong and irrelevant conclusions.

    You must be really depserate to resort to that paper. If that is your “damning experimental evidence”, which ” informs us that gpuccio is wrong”, then gpuccio must really be right.

    But why am I losing time with you? Your bad faith and ignorance have already been clearly shown.

    If you want to go on discussing, please before:

    a) Admit your gross error in dividing exponential values by a linear measure of time. You have not yet done that. What are you, a person without any dignity?

    b) Please explain what do you mean when you speak of “Starting with a transcription factor that binds DNA” about the Szostak paper. I really want to understand.

  230. 230
    gpuccio says:

    Corey Delvine:

    “Or is he arguing that there’s no potential for function in ATP-binding?”

    I hate to remind you how neo-darwinism and natural selection work: they are not interested in “potentials for function”. Not at all.

    NS is only interested in existing reproductive advantages.

    Design, instead, is often interested in “potential for function”. That’s how design works, not how NS works.

    Do you want an advice? Try to understand what you believe in. It can be useful, you know.

  231. 231
    gpuccio says:

    Corey Delvine:

    Now I have to sleep. If you have other brilliant comments, I will answer them tomorrow.

    But again, answer those two points:

    a) Admit your gross error in dividing exponential values by a linear measure of time. You have not yet done that. What are you, a person without any dignity?

    b) Please explain what do you mean when you speak of “Starting with a transcription factor that binds DNA” about the Szostak paper. I really want to understand.

  232. 232
    Corey Delvine says:

    Cho & Szostak 2006 starts with retinoid-X-receptor, a transcription factor, apparently I was referring to a different paper by Szostak in comment 226, but it changes nothing about what I said.

    Ah, so your objection to the Szostak paper is based on claiming that ATP-binding is a “useless function,” got it.

    Just so there is no confusion, you expect us all to believe that binding what is probably the single most important molecule at the cellular level is “useless?”
    Or is that just what you have to continuously tell yourself so that you can believe your number crunching pipe dream?

    Once again, who do we believe?
    Gpuccio, who is claiming that the experiment has nothing to do with natural selection and protein sequence space,
    or the scientists who actually performed the experiments?

    Oh, now you resort to harping on my misunderstanding of your ridiculous claims and demand that I admit my errors?
    Yeah, my toddler throws fits too.

    According to you, the only way to properly “simulate natural selection” is to create an entire ecosystem and monitor every single aspect for millions of years to get a step-by-step, molecule-by-molecule progression of evolution.

    Well, who needs to move the goalposts when you already have them set somewhere completely unattainable?
    You truly are the poster child for confirmation bias Gpuccio.

  233. 233
    ET says:

    Corey- Whoever is making the claim has to support it. That means if someone is saying that an ATP binding protein evolved by natural selection they have to back that up. And that would mean explaining where they got the ATP from. Transcription factor? Just how does natural selection explain their existence?

  234. 234
    Corey Delvine says:

    Wow ET, I told myself I wouldn’t reply to your nonsense, but this one takes the cake.

    So backing up the claim “that an ATP binding protein evolved” requires “explaining where they got the ATP from” as well?

    If gpuccio is the poster child for confirmation bias, your are the poster child for goalpost-moving.

    You guys are really covering all your bases!

  235. 235
    gpuccio says:

    Corey Delvine:

    OK, you have shown who you are. I don’t think you deserve any more attention. Good luck.

  236. 236
    Corey Delvine says:

    Good evening pucci!

  237. 237
    gpuccio says:

    To all interested:

    The Cho and Szostak paper of 2006 is the following:

    Directed Evolution of ATP Binding Proteins
    from a Zinc Finger Domain by Using mRNA Display”

    Emphasis mine.

    It is explicitly a paper about protein engineering, and not about natural evolution. IOWs, it makes clear what the Keefe and Szostak paper of 2001 kept hidden.

    Here is the abstract:

    Antibodies have traditionally been used for isolating affinity reagents to new molecular targets, but alternative protein scaffolds are increasingly being used for the directed evolution of proteins with novel molecular recognition properties. We have designed a combinatorial library based on the DNA binding domain of the human retinoid-X-receptor (hRXRa). We chose this domain because of its small size, stable fold, and two closely juxtaposed recognition loops. We replaced the two loops with segments of random amino acids, and used mRNA display to isolate variants that specifically recognize adenosine triphosphate (ATP), demonstrating a significant alteration of the function of this protein domain from DNA binding to ATP recognition. Many novel independent sequences were recovered with moderate affinity and high specificity for ATP, validating this scaffold for the generation of functional molecules.

    IOWs, they used the original TF to have a pre-existing stable fold (which was probably too difficult to obtain by their protein engineering), and then they artificially “evolved” ATP binding in two small parts of it.

    The first zinc finger loop was replaced with twelve random amino acid positions, while nine random positions were substituted for the second zinc finger loop. The finished library has a length of 96 amino acids, with 21 randomized positions

    So they obtained a rather big starting random library:
    The final complexity is estimated as 1.4 x 10^14 different molecules

    The 20 ATP binding sequences they obtained after 8 rounds of mutation and artificial selection by ATP-derivatized agarose beads share 4 AA positions.

    4 conserved AAs. About 16 bits of functional information for the simpe (and biologically useless in itself) function of ATP binding.

    So, to sum it up, they could generate 16 bits of functional information for ATP binding by artificial protein engineering helped by the utilization of an existing natural fold, and starting with a library of 1.4 x 10^14 different molecules.

    At least, there is no confusion here about the true purpose of the study. The authors state explicitly:

    We are interested in extending the discovery of new functional molecules from scaffolds by using mRNA display

    IOWs, they are interested in artificial protein engineering. And that is perfectly fine.

  238. 238
    gpuccio says:

    To all interested:

    Of course, none of the two quoted papers by Szostak has anything to do with NS. The first one remains ambiguous about that, the second one is more explicit and declares just from the beginning that it is a paper about artificial protein engineering.

    None of them is a simulation of NS.

    And, of course, it is perfectly possible to make a correct simulation of NS in the lab.

    That’s exactly what Hayashi et al. did in the famous rugged landscape experiment.

    How did they do such a thing? It’s simple.

    They slelected for infectivity, which in phages is the equivalent of reproductive success.

    That is a simulation of NS.

    With the results we have discussed in this thread, in detail.

  239. 239
    Corey Delvine says:

    Ah yes, there’s the famous spin doctor gpuccio at his best, explaining away any experimental evidence that contradicts his pipe dream science.

    “Szostak didn’t create an entire ecosystem in the lab in order to properly simulate natural selection”

    “The word ‘directed’ is in the title, therefore the entire paper has nothing to do with natural selection and it is actually protein engineering”

    Yes, thank you gpuccio, we really needed you to clear things up for us, otherwise we may have actually realized just how ridiculous your claims are.

  240. 240
    gpuccio says:

    To all interested:

    Well, Corey Delvine has abandoned any pretense of reasonable discussion, and is resorting to pure name-calling: I must have touched some sensitive spots.

    While that is certainly a very good result for ID theory, for some strange reason I am never really happy with that kind of ending.

  241. 241
    gpuccio says:

    To all interested:

    Is there any serious interlocutor left in the opposite field, to discuss with?

    Almost 140 comments ago, I made a challenge. Nobody has even tried to answer it.

    While each individual discussant in the neo-darwinian field is obviously absolutely free to decide if he should answer such a challenge or not, the fact that collectively nobody has tried is somewhat telling.

  242. 242
    gpuccio says:

    To all interested:

    By the way, I really want to thank again Gordon Davisson.

    While he has apparently retreated from the discussion, and has not considered my challenge (which, of course, he is perfectly entitled to), he has given great contributions here.

    And a very good example of intelligence and intellectual honesty.

    Thank you again.

  243. 243
    gpuccio says:

    To all interested:

    By the way, the “Clarke et al.2007” paper quoted by Corey Delvine at #226 is the following:

    “The folding and evolution of multidomain proteins”

    https://www.nature.com/nrm/journal/v8/n4/full/nrm2144.html

    And here we have, again, a good example of quote-mining, out of bad faith or simply ignorance.

    Corey states the following about that paper:

    This paper looked at 1.1 million protein sequences and found that when taking both sequence and structure into account:
    “consistently isolated superfamilies are rare” and “isolated domain superfamilies occur in only 0.16% of protein sequences”

    So, the vast majority of proteins are made up of similar folds?

    Ok, one explanation is that the amount of space evolution actually searches is very small compared to the potential space that it can search and this is due to evolution occurring in a stepwise fashion from existing domains (hence the similarity in all proteins).

    Emphasis mine.

    But here is the abstract of the paper:

    Analyses of genomes show that more than 70% of eukaryotic proteins are composed of multiple domains. However, most studies of protein folding focus on individual domains and do not consider how interactions between domains might affect folding. Here, we address this by analysing the three-dimensional structures of multidomain proteins that have been characterized experimentally and observe that where the interface is small and loosely packed, or unstructured, the folding of the domains is independent. Furthermore, recent studies indicate that multidomain proteins have evolved mechanisms to minimize the problems of interdomain misfolding.

    And here is the full quote of the part mentioned by Corey:

    Genomic structural assignments.
    Using the SUPERFAMILY database (version 1.69; see Further information) (BOX 2), we analysed protein domain architectures of 358 completely sequenced genomes from all three kingdoms of life (encompassing nearly 2 million protein sequences). Structural domains could be assigned to about 1.1 million of these protein sequences, approximately 65% of which contained more than one domain. Considering gaps (that is, regions of sequence with no assignments) as a single domain, about 95% of multidomain proteins contain 2–5 domains. A few proteins are comprised of many more (up to 303) domains. From our structural assignments, the following general characteristics are observed.
    Consistently isolated superfamilies are rare. Over the course of evolution, domains have undergone extensive recombination, and isolated domain superfamilies that never recombine with others are rare. Our analysis shows that only 86 out of 1,439 domain superfamilies are consistently found as single-domain proteins. These isolated domain superfamilies only occurred in ~1,700 protein sequences (0.16%) in our analysis.

    Emphasis mine.

    The meaning here is completely different from what Corey states.

    It is not true that “the vast majority of proteins are made up of similar folds”.

    And there is no “similarity in all proteins”.

    “Consistently isolated superfamilies”, as is clear from the full quote, are those superfamilies whose domains are never found in multi-domain proteins. That is very clear from the following phrase:

    “Our analysis shows that only 86 out of 1,439 domain superfamilies are consistently found as single-domain proteins.”

    IOWs, almost all domain superfamilies contribute to multi-domain proteins. Only a few of them are never found in multi-domain proteins.

    That has nothing to do with “the vast majority of proteins being made up of similar folds”, least of all with any “similarity in all proteins”. And the related ramblings and pseudo-reasonings.

    Again, gross ignorance or gross bad faith. You choose.

  244. 244
    Origenes says:

    No sane Darwinian dares to accept the challenge?

    Almost 140 comments ago, I made a challenge. Nobody has even tried to answer it.

    While each individual discussant in the neo-darwinian field is obviously absolutely free to decide if he should answer such a challenge or not, the fact that collectively nobody has tried is somewhat telling.
    [GPuccio]

    I nominate “somewhat telling” for The Understatement of the Year Award.

  245. 245
    gpuccio says:

    Origenes:

    “I nominate “somewhat telling” for The Understatement of the Year Award.”

    Thank you! I like awards. 🙂

  246. 246
    Mung says:

    Anyone else think “Corey Delvine” is a sock puppet?

  247. 247
    gpuccio says:

    Mung:

    You are probably right.

  248. 248
    ET says:

    Corey:

    So backing up the claim “that an ATP binding protein evolved” requires “explaining where they got the ATP from” as well?

    Well, yeah. You don’t get to start with the very things that need explaining. And again ID is not anti-evolution. The claim is that blind and mindless processes cannot produce proteins from scratch, ie without any existing proteins to help and without any existing designed molecules to choose from.

    But then again you seem to be too clueless to understand that.

  249. 249
    Dionisio says:

    Origenes @244 (in reference to gpuccio @241):

    I nominate “somewhat telling” for The Understatement of the Year Award.

    I second the motion for the nomination.

  250. 250
    ET says:

    Directed evolution of ATP binding proteins from a zinc finger domain by using mRNA display. Cho, Szostak, used DIRECTED evolution. That is not only in the title but also prevalent in the paper.

    ID argues against blind and mindless processes producing proteins in the first place. And if you want to believe that blind and mindless processes can produce other proteins given starting organisms complete with proteins, then you have to address the paper waiting for two mutations. This is because it is obvious that more than two specific mutations would be required to do so.

    But then again Corey will never be able to understand any of that. Willful ignorance and arrogance are Corey’s “weapons”.

  251. 251
    Dionisio says:

    Perhaps the politely dissenting interlocutors should get partial credits for this stats?

    Popular Posts (Last 30 Days)

    Violence is Inherent in Atheist Politics (1,997)

    What are the limits of Natural Selection? An interesting… (1,872) [since October 5]

    Howling Darwinists (1,470)

    Selective Horrid Doubt (1,316)

    Sweeping the Origin of Life Under the Rug (1,001)

    🙂

  252. 252
    Dionisio says:

    ET @250:

    Willful ignorance is the worst kind, because nothing can reduce it. Actually, nothing can keep it from getting worse.

    But there’s hope. I was there myself years ago.
    The same God they hate, loves them more than they can understand. God’s grace can make miracles, I’m a living proof.

  253. 253
    Origenes says:

    GPuccio @243 : Again, gross ignorance or gross bad faith. You choose.

    I go with gross ignorance. I mean, Corey Delvin is a guy who believes that natural selection in the lab requires building “an entire ecosystem and monitor every single aspect for millions of years” (post #228, #232 and #239). This is not at all a joke by Corey, he seriously holds that when you are talking about artificial selection that you are objecting to the absence of such an ecosystem. That is his genuine understanding of the discussion….
    No one has commented on this, because it is simply stupid beyond belief.

  254. 254
    gpuccio says:

    Origenes:

    “No one has commented on this, because it is simply stupid beyond belief.”

    True!

  255. 255
    gpuccio says:

    To all interested:

    I want to sum up some of the main conclusions that can be drawn from the interesting and multi-faceted discussion in this thread.

    The first and foremost derives from my challnge at #103, which has not yet been addressed by anyone in the opposite field.

    The important conclusion about that point can be expressed in a stronger form, or in a weaker form. Here they are:

    Stronger form:

    1a) There are no pathways that allow to build any complex function in proteins from simple naturally selectable steps.

    Weaker form:

    1b) There are no known conceptual reasons, and no known empirical facts, that support the idea that complex functions in proteins can be deconstructed into simple naturally selectable steps.

    Whatever form you like, the result is the same: NS cannot be considered a credible scientific explanation for complex functions in proteins.

    That is the simple truth.

    This conclusion remains, IMO, the single most striking evidence against the role of NS in the imaginary neo-darwinian scenario. Not the only one, not at all, but certainly the most important.

    Of course, my challenge remains open to all. If anyone is aware of conceptual reasons or empirical facts that support that strange idea, they can speak at any time.

    More in next post.

  256. 256
    gpuccio says:

    To all interested:

    The second important point regards the powers of Random Variation. It can be summarized as follows:

    2) Random variation, of any kind, is completely powerless in regards to the task of generating any new complex functional information.

    In my post #209, I have argued in great details why that is the case. I invite all those who are interested to read that quantitative argument.

    This conclusion applies to all forms of random variation: single nucleotide mutations, translocations, duplications, inversions, insertions, deletions, recombination, you name it. Including fixation by genetic drift, which is a random event like all the rest.

    In all forms of RV, each variation has the same meaning: a new state in the search space is tested.

    While a local search has certainly greater power in the case of tweaking an existing function, and recombination has greater power in remixing existing information in some new form, in the general case where new complex functional information comes into existence all forms of RV are equivalent: they are steps of a random walk to a functional target starting from some unrelated sequence, and each step, whatever it is, can only test one state. The probabilities remain the same for all forms of RV.

    So, our conclusions at point 1 and 2 imply the following important conclusion:

    The neo darwinian scenario relies only on two components to explain new complex functional information: RV and NS. Nothing else.

    As RV is completely incapable of generating new complex functional information, and NS is equally powerless, because no natuarally selectable pathway exists to generate new complex functional information by simple steps in the range of RV, then the simple truth is that:

    3) The theory based on RV and NS (neo-darwinism) is absolutely incapable to explain how new complex functionl information appears in natural history.

  257. 257
    gpuccio says:

    To all interested:

    Now, someone could ask:

    But does complex functional information appear in natural history?

    Of course, the answer is yes!

    I invite all those who are interested to read my post #173 (the final part), where I give the distribution of the jump in human conserved information from pre-vertebrates to vertebrates. Here it is, again (in bits):

    Mean: 189.5817

    SD: 355.6362

    IQR: 224

    Median: 99

    90th percentile: 486 bits

    Number of proteins evaluated: 19564

    (I have corrected a typo in the original post)

    The value of the 90th percentile means, very simply, that 10% of human proteins have a jump in human conserved information higher than 486 bits, in that transition!

    That means almost 2000 proteins!

    So, we can conclude:

    4) Complex functional information does appear in natural history. It appears often, and in huge quantities, and often in rather short evolutionary times (in our case, about 30 million years).

  258. 258
    Origenes says:

    GPuccio #255, #256
    I would like to note that there are some serious hurdles further down the road.

    Suppose that, against all odds, RV produces a protein with function X. Is this reason for Darwinians to go celebrating? Not quite yet. In fact, not even close. Now the question arises, “does it fit the organism?”

    Not any old function fits. Obviously, an elephant has no use for a pair of wings.

    And even when there is a potential fit, there is a myriad of concerns such as: Is the new protein produced at the right time, in the right amount, at the proper place? Does it reach its proper destination? Is homeostasis not disturbed?

    So we have to ask ourselves: What are the chances that a new protein fits an organism?

  259. 259
    gpuccio says:

    Origenes:

    Of course you are right.

    That’s why, in all myb statements here, and in my challenge, I have explicitly used the form:

    naturally selectable steps

    Not function. Not vague fitness.

    NS requires a change that gives a real reproductive advantage versus the previous allele, to have a chance to act. Nothing else will do.

    IOWs, a positive selection coefficient is needed, and even that does not guarantee the fixation: there is always a stochastic aspect, and the competition with neutral drift. And the fixation time is another important aspect, because it can be extremely relevant.

    But again, why focus on those aspects, when the role of NS for complex functional information is, just from the beginning, completely inexistent?

    (see points 1 and 2, posts #255 and #256)

  260. 260
    Mung says:

    IOWs, a positive selection coefficient is needed, and even that does not guarantee the fixation: there is always a stochastic aspect, and the competition with neutral drift.

    That’s right. Even natural selection is stochastic. Random.

    So the next time someone claims evolution is not random you’ll know they don’t know what they are talking about.

  261. 261
    gpuccio says:

    Mung:

    “That’s right. Even natural selection is stochastic. Random.”

    I agree only in part.

    There is a random component in the eventual fixation of a naturally selectable trait, but while the fixation of a neutral trait is completely random (IOWs, all neutral traits have the same probability to be fixed, and the selection coefficient is zero), the fixation of a trait that gives some reproductive advantage is in part due to a necessity component.

    In that case, indeed, the selection coefficient is positive, and that means that, although the system remains in part stochastic, and it is always possible that some neutral trait be fixed instead of the selectable trait, IOWs that genetic drift prevail over NS, it is equally true that the probabilities to be fixed of the selectable trait are higher than the probabilities of any other neutral trait.

    Now, that “advantage” (expressed in the positive selection coefficient) is not a random component: it is connected to a specific property of the new allele, to the reproductive advantage that it has in the system. Therefore, that advantage is a result of necessity laws, and is not random.

  262. 262
    Origenes says:

    GPuccio @259

    naturally selectable steps …

    NS requires a change that gives a real reproductive advantage versus the previous allele, to have a chance to act. Nothing else will do.

    Including protein sequences with (unfitting) function and even including protein sequences with potential fitting function, but without proper regulation!

    But again, why focus on those aspects, when the role of NS for complex functional information is, just from the beginning, completely inexistent?

    Of course you are right. Why keep beating a dead horse?

  263. 263
    gpuccio says:

    Origenes:

    “Including protein sequences with (unfitting) function and even including protein sequences with potential fitting function, but without proper regulation!”

    Of course!

    For example, Szostak’s ATP binding protein is completely useless in a biological context, indeed deleterious.

    What is the utility of a protein that just binds ATP? It can only subtract precious ATP from the system!

    The whole idea of ATP is that it is a repository of chemical energy, to be spent for the various biochemical necessities of a cell.

    So, ATP synthase builds ATP from proton gradients (which, in turn, derive from cell metabolism), and othe rproteins use ATP to get the energy for other things.

    Let’s take, for example, this brief description of dynein from Wikipedia:

    Cytoplasmic dynein, which has a molecular mass of about 1.5 megadaltons (MDa), is a dimer of dimers, containing approximately twelve polypeptide subunits: two identical “heavy chains”, 520 kDa in mass, which contain the ATPase activity and are thus responsible for generating movement along the microtubule; two 74 kDa intermediate chains which are believed to anchor the dynein to its cargo; two 53–59 kDa light intermediate chains; and several light chains..

    In general, we have to have four different functions to do something useful with ATP:

    a) ATP binding

    b) ATPase activity, which releases chemical energy

    c) The rest of the molecule, which implements the real function

    d) An efficient coupling between b) and c), IOWs an efficient way to transfer the energy to the final function.

    So, in dynein, ATPase activity is coupled to the rest of the protein, whih uses the released energy to move along the microtubule, transporting the appropriate cargo to the appropriate target point.

    Of course, a designer can understand that, to design the whole complex, he needs first of all ATP binding.

    A designer can understand that. NS cannot.

    “Of course you are right. Why keep beating a dead horse?”

    That’s exactly my idea! 🙂

  264. 264
    Dionisio says:

    Popular Posts (Last 30 Days)

    Violence is Inherent in Atheist Politics (2,005)

    What are the limits of Natural Selection? An interesting… (2,001) [since October 5]

    Howling Darwinists (1,472)

    Selective Horrid Doubt (1,316)

    Sweeping the Origin of Life Under the Rug (1,001)

    The NS topic seems to attract attention.

  265. 265
    Dionisio says:

    Origenes @258:

    Obviously, an elephant has no use for a pair of wings.

    Are you sure?

    http://es.web.img2.acsta.net/p.....045404.jpg

    🙂

  266. 266
    gpuccio says:

    Dionisio:

    Ah, Dumbo! 🙂

  267. 267
    gpuccio says:

    Dionisio:

    That could be an interesting conflation of two fundamental themes: the evolution of the ear and the evolution of wings! 🙂

    Covergent evolution, again?

  268. 268
    Dionisio says:

    gpuccio,

    That’s an interesting research topic: wingear which perhaps eventually evolved into the ‘win gear’ – the mechanism Neo-Darwinian folks use to always win the discussions.

    Wing ear —> wingear —> win gear!

    🙂

  269. 269
    Dionisio says:

    Popular Posts (Last 30 Days)

    What are the limits of Natural Selection? An interesting… (2,023) [since October 5]

    Violence is Inherent in Atheist Politics (2,007)

    Howling Darwinists (1,473)

    Selective Horrid Doubt (1,316)

    Sweeping the Origin of Life Under the Rug (1,001)

  270. 270
    Mung says:

    gpuccio:

    I agree only in part.

    There is a random component in the eventual fixation of a naturally selectable trait, but while the fixation of a neutral trait is completely random (IOWs, all neutral traits have the same probability to be fixed, and the selection coefficient is zero), the fixation of a trait that gives some reproductive advantage is in part due to a necessity component.

    I’d like you to re-read what you wrote and note how you switched terms in mid stream. 🙂

    You are willing to say that fixation of a neutral trait is “completely random”, but don’t seem willing to say that fixation of a non-neutral trait is “partly random.” But it most certainly is, at a minimum, partly random. Wouldn’t you agree?

    Instead you insist that fixation of a non-neutral trait is due to some mysterious “necessity component.” And here you have lost me.

    Say you have many traits in the population with equal selective values. Do they not also all have the same probability to be fixed, just as in the neutral case?

    I don’t mean THE SAME probability as in the neutral case. I mean equal probability among the cases that share the same selective value.

    Will all of them be fixed? So we can say using your definition that since they all have the same probability, they are “completely random.”

    There is certainly a probability that not all will not be fixed. Don’t you agree?

    So what “necessity component”?

    Both neutral and non-neutral spread is probabilistic. They are both stochastic. Both are random. Unless we are going to redefine stochastic.

    Much respect sir! Hope this makes you think.

    It is not Chance and Necessity as Monod claimed. It is Chance and Chance. With some having better chances than others, lol.

  271. 271
    gpuccio says:

    Mung:

    First of all, I want to say that I have for you the greatest esteem (and a huge appreciation for your sense of humour!)

    Moreover, I am certain that we agree on all important things.

    For those reasons, I think that it is important for me to clarify my views on this point, about which we seem to differ. Of course, if after all the necessary clarifications we still have different ideas, there’s no problem at all. 🙂

    I say that we “seem to differ” because, after readin carefully what you said in #270, I believe that part of our “divergence” is due only to a problem of words. But part of it could still be a true difference. Let’s see.

    A necessary premise to all the following discussion is that all the concepts I will discuss are always related to models of reality: our models, and in particula our scientific modles. That’s why the word “model” will recur often.

    The first problem is that we seem to use the word “necessity” in a different way. There is no surprise in that: the word itself is many-faceted, in philosophy and in science. But words are not really a problem, if we clarify what we mean with them.

    From what you say, I understand (please, correct me if I am wrong) that you use the word in a perfectly correct, but rather strict, sense. For the sake of clarity, I will call that sense of the word, at least in this comment:

    “absolute necessity”.

    I will also try to define it explicitly, so that there may be no confusion.

    Let’s say that we have a model that includes two variables, A and B. Let’s also remember that we are discussing empirical models, not pure logic. So, A and B are observable things. Facts.

    Let’s say that absolute necessity means a model where, if A happens, B must happen (or not happen, which is the same). We could also say that, if A happens, the probability that B happens is 1 (or 0, in the opposite case).

    In most cases, we interpret that kind of observation, if all the methodological cautions are well satisfied, as a causal relationship between A and B: we say that A is the cause of B.

    For the sake of this discussion, we can assume that we find that kind of scenario in many contexts of physics: that is not completely true, but true enough for our discussion.

    So, the law of gravitation according to Newton states that the masses of two objects and their distance are the cause of the gravitational attraction between two bodies, or at least explain that attraction very well, according to a precise mathematical relationship.

    A (the two masses and the distance) explains B (the gravitational attraction).

    As that relationship can be observed easily, in practically all contexts, and with great precision, we call that (supposedly) causal relationship a law. In particular, a law of necessity, because, given the right masses and distance, the probability of having a specific force of attraction is practically 1. Especially if we can control well, in our experiemnts or observations, all disturbing variables that could interfere.

    Let’s call a model that describes (well enough) some physical system using only strict necessity a fully deterministic model.

    OK with that?

    So, in the laws of physics, or at least in some of them, we can find a very good approximation of the concept of empirical strict necessity. Absolute necessity, according to my conventional name here.

    Now, what happens in other sciences, like biology or medicine?

    In those fields, strict necessity in that sense is really a rare thing. The systems we are considering are too complex, the variables are too many, and of many of them we are not aware, or cannot measure any value.

    That’s why we use, in almost all cases, probabilistic models to describe biological reality and biological data.

    That means that we cannot speak any more of absolute necessity. Does that mean that we cannot say aything about laws in biology? Does that mean that biology is the kingdom of chance alone?

    Not at all.

    Let’s say, again that we have A and B. But now A and B are biological data, in particular biological variables.

    For the reasons I have said, both those variables behave as random variables, IOWs variables that can assume different values, and whose value distribution can be described with some appropriate probability distribution.

    Let’s say, for example, that A is the set of values of the blood levels of growth hormon at some age (for example, at 5 years) in some population, and B is the final height of those individuals.

    Both A and B can be considered random variables, because both can be well described by some probability distribution, in this case probably the normal distribution.

    But science does not stop there. We ask ourselves specific questions, like the following:

    Is there any causal relationship between the values of A and the values of B?

    To assess that, we need to perform a statistical analysis, in the correct methodological context.

    From a statistical point of view, we need to assess if the two variables are independent, or if there is some specific relationship between them.

    Well, I will continue in my next post.

  272. 272
    gpuccio says:

    Mung:

    Now, if the two variables are really independent, then knowing the value of A will have no effect on the probability of B. IOWs, the levels of growth hormone will have no relationship at all with the final height.

    But that is probably not the case. Children with low levels of growth hormone will have low probabilities of being tall. That means that there is a statistical relationshipb between the two variables, what we call an effect. Which, in the right methodological context, can be intepreted as a causal relationship between levels of growth hormone and final height.

    So, would it be correct to say that height is due to mere chance?

    No.

    And yet, our model is probabilstic.

    Let’s make another example.

    A is exposition of a population of individuals to influenza virus: some of them have been in contact with people with the disease, some have not.

    B is the probability of developing respiratory symptoms in the following 3 days. Some will, some will not.

    Of course, not all exposed people will developed respiratoy symptoms, and not all non exposed people will not develop them.

    So, is there a relationship of strict, absolute necessity here?

    No. Not in your sense of the word. The probability of developing symptoms in exposed people is not 1.

    We have a probabilistic model, again.

    But can we say that the development of respiratory symptoms is due to chance alone?

    No, we can’t. Exposed people are much more likely to develop symptoms.

    That’s what I call “a neccessity component” in the system. Exposure to influenza is a cause of respiratoy symptoms. Not an absolute cause, and not the only cause. But it is a cause.

    In physicis, we had that:

    A causes B, with probability (practically) 1.

    We called that a law.

    Now, in biology, we have that:

    A causes a detectable variation in the probability distribution of B.

    That is a law, too. Or at least a definite causal relationship, which expresses itself as a definite, detectable effect on the probabiltiy distribution of B.

    So, my point is that in that situation we cannot say that we are in front of “chance alone”. We are in front of chance + detectable causes. If you don’t want to call that “necessity”, or, as I did, “a necessity component”, I have no problems with that. We can simply call it “a causal relationship”.

    But I don’t believe that we can call it “chance alone”.

    Finally, does that scenario apply to what we were discussing, in particular NS?

    I believe it does.

    A, here, if the nature of the random variation, as assessed by its effects of reproduction. It can be expressed in categorical form (deleterious, neutral, beneficial), or in continuous form, as a selection coefficient.

    B is the probability of fixation.

    Again, the system is probabilistic. Not all beneficial variations are fixed, and many neutral variations are fixed by genetic drift, even in competition with NS.

    So, can we say that chance alone is acting in the system?

    No, because we can find a statistically detectable relationship between A and B: variations with a positive selection coefficient have higher probability to be fixed, and the higher the coefficient, the higher the probability.

    It’s not chance alone. It’s a causal relationship. A probabilistic law.

    That’s what I meant by “a necessity component”. However we decide to call it.

  273. 273
    Mung says:

    Very fine posts gpuccio! Thank you!

    We do agree on so much.

    So if you have an urn containing 20 red marbles, 30 blue marbles and 50 green marbles and you draw a marble “at random” from the urn, what other force or cause is operating other than “chance alone” when it comes to the color of marble that is drawn from the urn?

    So say you draw a green marble. The chance of a green vs a not green is 50/50. Why not call that “chance alone.”

    🙂

  274. 274
    gpuccio says:

    Mung:

    In the situation you describe, the system is completely random, but the probability distribution that describes the system is not uniform. That is not a problem, uniform distribution is only one of the probability distributions that describe physical systems.

    Green marbles have simply a higher probability in the distribution, because there are more of them in the system. Of course, the sum of all probabilities, in discrete distributions, must be 1.

    But here we cannot gain any further knowledge about the probabilities of the draws from another variable. There is no causal relationship between two variables, therefore no scenario where, knowing the value of A, you gain additional information about probabilities in B.

    For example, let’s say that some disease, in a population, has a prevalence of 3%. That means that, if you draw by chance some sample from the population, you will have approximately 3% of the sample with the disease (and that will be more precise as the size of the sample increases).

    Now, let’s say that we have another categorical variable which can be observed in the same system, for example sex.

    Let’s say that 60% of the population is male, and 40% of the population is female.

    So, the distribution of the sex variable is not uniform in the population.

    Now, there are two possibilities:

    a) The sex variable and the disease variable are independent.

    That means that not only 3% of the population in general has the disease, but also 3% of the male population and 3% of the female population.

    So, knowing if an individual is male of remale does not change his/her probability of having the disease.

    There is no relationship between the two variables.

    b) While the prevalence of the disease is 3% in the whole population, its prevalence is higher in males. Let’s say that it is 4% in males, and 1.5% in females. With the male/female ratio we have given, that would result in exactly 3% of the disease in the general population.

    But now, if we draw a sample from the male population, we have a probability of disease which is about 2.66 times higher than if we draw a sample from the female population.

    Therefore, if we know in advance if an individual is male or female, we have added information about his/her probability of having the disease.

    Why is that?

    Because the relationship between the categorical variable sex and the categorical variable disease is not a relationship of independency. The two variables are dependent.

    And, as the sex variable is probably established before the onset of the disease, the best explanation for that scenario is that being male has a causal role in favouring the disease.

    Therefore, the distribution of the variable “disease” is not random in relation to the variable sex: it is influenced by it, by a causal relationship.

    Of course, the variable “disease” can at the same time be completely random in relation to other variables in the system: for example, race. Or race can also have a causal relationship with the disease. Both scenarios are possible. Only a correct statistical analysis of the observed facts can tell us what the best explanation is, and the level of confidence we can have in our conclusions.

    I hope that helps.

  275. 275
    gpuccio says:

    To all interested:

    There are indeed a few aspects about NS that probably have not been touche in detail in our discussion. So, I would like to say something about them.

    Here is the first:

    a) The differences between Natural Selecion (NS) and Artificail Seleciton (AS, aka Intelligent Selection, IS).

    I have dedicated a whole OP to this issue:

    https://uncommondescent.com/intelligent-design/natural-selection-vs-artificial-selection/

    However, I would like to summarize her the main differences, and add a few comments.

    First of all I paste here the final conclusions of my OP:

    1) AS can define any function, and select for it. NS works only on one function: reproductive success.

    2) In NS, the coupling between function and selection is direct: it’s the function itself which confers the reproductive advantage, which is the reason for the selection itself. In AS, the coupling between the defined function and the selection process is indirect and symbolic: the connection is established by the designer, by definite procedures designed by him.

    3) NS has a definite threshold of measurement: it can only act if enough reproductive success is present as to ensure the fixation of the trait. AS can measure and select any desired level of the defined function.

    4) In NS, the only selecting procedure is tied to the reproductive success, and is in essence differential reproduction. In AS, any intelligent procedure can be used to isolate, expand and fix the desired function.

    That said, I would like to emphasize a special aspect of the issue:

    NS is a subset of the general set of Selections (possible forms of selection, S). Indeed, an extremely small subset.

    In fact, NS can be defined as a form of S where:

    1) The only defined function is reproductive success in some system.

    2) The coupling between function and selection is implicit in the defined function itself, and need not be implemented explicitly in the system in some indirect and symbolic way.

    3) The measurement of function is implicit in the defined function, and hase a somewhat fixed threshold (the function must be strong enough to give a detectable reproductive advantage, expressed as a relevant selection coefficient).

    4) The selecting procedure is, again, implicit in the defined function, and needs not be implemented in the system.

    As anyone can see, these specific criteria are hugely restrictive.

    So, if we imagine the set of all possible forms of selection, NS will be an extremely tiny subset. All the rest will be forms of AS.

    This explains the different power of NS and AS in generating functional information. I will say more about that later.

  276. 276
    Mung says:

    gpuccio:

    In the situation you describe, the system is completely random, but the probability distribution that describes the system is not uniform. That is not a problem, uniform distribution is only one of the probability distributions that describe physical systems.

    Green marbles have simply a higher probability in the distribution, because there are more of them in the system. Of course, the sum of all probabilities, in discrete distributions, must be 1.

    How does this differ from natural selection?

  277. 277
    gpuccio says:

    To all interested:

    b) Can NS be simulated in the lab?

    I have alredy answered that question in the previous discussion, but I would like to emphasize some points.

    The answer is: yes, but the simulation must be formally appropriate, if we want to draw realistic concusions about how NS works in the wild.

    To clarify that point, I have compared two different experiment:

    1) Szostak’s experiment about the generation of ATP binding sequences.

    2) Hayashi’s experiment about the rugged landscape.

    Both are lab simulation. Both seem to be, more or less explicitly, about what NS can do.

    But my simple conclusion is that:

    1. is not a simulation of NS at all. It is only an example of AS.

    2. is an appropriate simulation of NS, form which we can draw some cautious but realistic conclusions about how it works.

    Why is that the case?

    It’s simple.

    In Szostak’s work, all of the properties of NS are lacking:

    1) The defined function is not reproductive success.

    2) The coupling between function and selection is not implicit in the defined function: it is indirect and symbolic: the connection is established by the specific settings of the experiment.

    3) The measurement of function is not implicit in the defined function: it is realized by columns of ATP-derivatized agarose beads, which are certainly not available in a natural scenario. Moreover the only lower threshold of detectable function was the limit of sensitivity of that technique, which is indeed capable of detecting even very low levels of ATP binding. We have also discussed the important concept that ATP binding, in itself, is not a naturally selectable function. I will say something more about that in a later post.

    4) The selecting procedure is completely artificial, and depends on the specific procedure used in the experiment: the sequences are isolated by the ATP binding columns, as decribed at 3), and then expanded by PCR or mutagenic PCR.

    Therefore, the procedure used by Szostak is not a simulation of NS: a simulation is certainly different from what happens in the wild for some aspects, but must retain some basic similarities to the process it is trying to simulate, if any valid inferences are to be drawn from it about the process itself. The Szostak procedure is about AS, and has none of the features of NS. It has nothing to do with NS, and certainly it is no simulation of it.

    The Hayashi experiment, instead, is bvery different. Let’s see:

    1) The only defined function is reproductive success, which is expressed here as infectivity. Indeed, for phages, the two concepts are practically the same thing.

    2) and 3) and 4) The coupling between function and selection, the measurement of the function and the expansion of selected sequences are not implicit: they are realized by the system. However, what is selected and measured is infectivity, that is reproductive success. And the sequences that are expanded are thjose with higher infectivity.

    So, what can we say? It is of course an artificial procedure (a simulation). Steps 2, 3 and 4 are simulated, and of course they do not happen like they would in the wild.

    However, we can assume that the general form of the process has some good resemblance with what would happen in the wild, because we are measuring, selecting and expanding infectivity, reproductive success. And that is what is supposed to happen in the wild, too.

    The measurement itself is a good simulation of NS, because the increase in infectivity must be detectable as nuber of infected colonies, which guarantees that such a level could probably be detected by NS in the wild.

    So, this is a good simulation of NS: some important aspects are simulated, but there are good reasons to think that the form of the observed process has good similarities to real NS, and that therefore good and valid inferences about NS can be drawn from the results.

    The important point, again is: they are using in their simulation the same property (reproductive success) which is the object of NS in the wild. Therefore, this is a valid simultaion, and it has relevance. Of course, it is not necessarily perfect, and has potential flaws, but at least it makes sense!

    The final conclusion is: NS can be simulated in the lab, but the simulation must make sense: the basci requirement is that only reproductive success can be used as a selectable property in the simulation. Anything else will be a simulation of AS, or simply an example of AS, and will have no connextion with NS.

  278. 278
    gpuccio says:

    Mung:

    “How does this differ from natural selection?”

    It is completely different.

    In your example, the color of the marble is related to the probability of being drawn simply because of the relative frequency of each color in the population. If the drawing is really random, then the relative frequency will be the only factor in fixing the probability of drawing some specific color. Each individual marble has the same probability of being drawn: as there are more green marbles, the color green has more probabilities of being drawn vs other colors. OK?

    Now, let’s try to apply your example to mutations. Let’s say that in a population you have 2997 neutral mutations, and 3 beneficial mutations, at some moment.

    We can ask: what is the probability of fixation of a beneficial mutation vs the probability of a neutral mutation?

    If the probability of each mutation to be fixed is the same, independently from the type of mutation, the general probability of having a beneficial mutation fixed will be 1:1000. IOWs, much lower than having a neutral mutation fixed. There is no selection for type of mutation.

    In this case, we are in the same situation as in your example with colored marbles: the probability of each color to be drawn depends only on the relative frequencies of each color, because color itself has no effect on the probability of being drawn. There is no selection for colors.

    But, if each beneficial mutation has a greater probability to be fixed than each neutral mutation, then the scenario is different.

    Let’s say that beneficial mutations have a probability of being fixed of 2:1000, which is still a low probability.

    That will make the probability of each neutral mutation to be fixed only slightly lower: 0.999:1000.

    However, now the probability of each beneficial mutation to be fixed is more than twice the probability of each neutral mutation.

    Here, the two variables are no more independent. The nature of the mutation influences the probability of each mutation to be fixed, while in the first scenario the probability was the same for each mutation, whatever the type, and the final probability of a beneficial mutation to be fixed depended only on the number of beneficial mutations available.

    In the second scenario, the probability of each type of mutation to be fixed always depends (obviously) on the number of available mutations of that type (that is always true), but it also depends on the type of mutation: if we refer to one single mutation, the probability of being fixed will be more than double if it is a beneficial mutation.

    Here the two variables are dependent, and there is a process of selection where the value of the first variable (not the number of available items) changes the probability of the second variable.

  279. 279
    Origenes says:

    GPuccio @272

    GPuccio: … variations with a positive selection coefficient have higher probability to be fixed, and the higher the coefficient, the higher the probability.
    It’s not chance alone. It’s a causal relationship. A probabilistic law.

    Can one determine the selection coefficient of a variety independent from the fixation in a population? I’m asking, because if there is no independent method to determine the selection coefficient — if the fixation in the population informs the selection coefficient — then I don’t see a valid basis for a law.

  280. 280
    gpuccio says:

    Origenes:

    “Can one determine the selection coefficient of a variety independent from the fixation in a population? I’m asking, because if there is no independent method to determine the selection coefficient — if the fixation in the population informs the selection coefficient — then I don’t see a valid basis for a law.”

    As usual, you ask a very good question. I had wondered about that too, while I was writing my last comments.

    OK, I am not really an expert in population genetics, but here is what I think.

    In general, the selection coefficient is probably derived indirectly from observations of the fixation in some particular case. For example, here is a brief statement in Wikipedia at the “selection coefficient” page:

    “For example, the lactose-tolerant allele spread from very low frequencies to high frequencies in less than 9000 years since farming with an estimated selection coefficient of 0.09-0.19 for a Scandinavian population. Though this selection coefficient might seem like a very small number, over evolutionary time, the favored alleles accumulate in the population and become more and more common, potentially reaching fixation.”

    And there is a reference to a paper:

    Bersaglieri, T. et al. Genetic signatures of strong recent positive selection at the lactase gene. Am. J. Hum. Genet. 74,1111-1120(2004).

    Moreover, I think that the Hayashi paper clearly shows that, in some scenarios, we can certainly measure directly the reproductive advantage of some mutational event (in that particular case by measuring infectivity). Measuring the reproductive advantage should be related to measuring directly the selection coefficient, although I could not say exactly how.

    Moreover, we have the few classic clear examples of microevolution that demonstrate how in extreme scenarios some simple mutation, if it can give an extreme reproductive advantage, is quickly fixed. In that case, we can certainly assume that the selection coefficient is very high, and the probability of fixation is near to 1.

    That’s the case, for example, for simple antibiotic resistance. It happens all the time, both in the wild and in the lab.

    So, we can know for certain that NS exists and that, in some specific cases, it works very well.

    That said, all the arguments about the limits of NS in this thread remain absolutely valid.

    Recognizing the few and simple things that NS can really do is one more strong argument against the imaginary things that it is supposed to do, and that it can never do.

  281. 281
    Origenes says:

    GPuccio
    Correct me if I am wrong, but I think that “selection coefficient” is synonymous with fitness — a notoriously troublesome term.

    Lewontin: A zebra having longer leg bones that enable it to run faster than other zebras will leave more offspring only if escape from predators is really the problem to be solved, if a slightly greater speed will really decrease the chance of being taken and if longer leg bones do not interfere with some other limiting physiological process. Lions may prey chiefly on old or injured zebras likely in any case to die soon, and it is not even clear that it is speed that limits the ability of lions to catch zebras. Greater speed may cost the zebra something in feeding efficiency, and if food rather than predation is limiting, a net selective disadvantage might result from solving the wrong problem. Finally, a longer bone might break more easily, or require greater developmental resources and metabolic energy to produce and maintain, or change the efficiency of the contraction of the attached muscles.

    and T. Dobzhansky wrote:

    … no biologist ‘can judge reliably which ‘characters’ are useful, neutral, or harmful in a given species.’

    [Quotes from this Barry Arrington article]

    While admittedly, sometimes mutations give a clear reproductive advantage, as you point out, but for most mutations this doesn’t seem to be the case. And I suspect that for those unclear cases the selection coefficient can only be determined by fixation rates — and not vice versa.
    And that seems to me to be identical with the same old tautology:
    Arrington: As we all know, Darwinian theory “predicts” that the “fittest” organisms will survive and leave more offspring. And what makes an organism “fit” under the theory? Why, the fact that it survived and left offspring.

  282. 282
    gpuccio says:

    Origenes:

    “While admittedly, sometimes mutations give a clear reproductive advantage, as you point out, but for most mutations this doesn’t seem to be the case. And I suspect that for those unclear cases the selection coefficient can only be determined by fixation rates — and not vice versa.”

    I can agree.

    My point is not that neo-darwinists are right in evaluating selections coefficients. I have no interest in that.

    My point is simply that, if a reproductive advantage really exists, like in the few cases I have mentioned, then NS can work.

    That’s why I have argued that the strongest argument against NS as a possible explanation for complex functional information is that complex functional information cannot be deconstructed into simple naturally selectable steps.

    From that, my challenge and all the rest.

    IOWs, NS can work in some very limited and simple cases, but is powerless in almost all the rest of what we observe in natural history.

    I believe that we cannot deny that NS works in simple and extreme cases of microevolution: that is the only small brick in the neo-darwinian castle which is supported by evidence.

    To deny it would be useless, indeed deleterious. It is simply true.

    And it is equally true that there is absolutely no support by evidence that NS can do anything more that that: explain very limited and simple cases of microevolution.

    If someone on the other side had any empirical support to the idea that complex functions can be deconstructed into naturally selectable steps, they would certainly have answered my challenge by now.

    Moreover, well before making my challenge here, I have been stating time and again in my comments here, for years, that complex functions cannot be deconstructed into simple naturally selectable steps. Nobody has ever been able to argue that it’s the other way round, or to show any evidence in that sense.

    But there is more: even artificial selection has severe limits, when acting on random variation, if new functional complex information has to be found. I will discuss that in my next post, maybe tomorrow.

  283. 283
    Origenes says:

    GPuccio @282

    GPuccio: But there is more: even artificial selection has severe limits, when acting on random variation, if new functional complex information has to be found. I will discuss that in my next post, maybe tomorrow.

    That would be crucial stuff, because, if artificial selection is severely limited, then, obviously, natural selection most certainly is.

  284. 284
    gpuccio says:

    To all interested:

    c) How is it that NS can generate functional information, even if in small quantities? Can AS do more, and how much more?

    Two questions, but strictly related.

    Some in the ID field find it difficult to believe that RV and NS can generate any functional information at all.

    But that is not suprising. In my old OP about functional information:

    https://uncommondescent.com/intelligent-design/functional-information-defined/

    I have said explicitly that we can accept any definition of function for the object we are observing. Any function will do, provided that we compute the complexity linked to it, IOWs, the minimal information that is required to implement that function.

    In that post, I have given the following example:

    I am a conscious observer. At the beach, I see various stones. In my consciousness, I represent the desire to use a stone as a chopping tool to obtain a specific result (to chop some kind of food). And I choose one particular stone which seems to be good for that.

    In this case, the function is very simple, and the specific information in the configuration of the object that we can use to implement it is very low: many stones will be good for the function.

    If I choose a stone, and I can perform the function with it, the stone has certainly the functional information required for that function. But the stone was not designed by anyone for that specific purpose: it is just one among many similar stones, the result of random natural forces.

    In the same way, if we consider, again, the Szostak experiment, he starts with some protein sequences from a random library that “naturally” (if we accept the random library as a given resource) have the property of buinding weakly to ATP (and therefore can be artificially selected by ATP binding columns).

    Now, we can certainly define “binding to ATP, even with a very weak binding” as a function.

    But, with that definition, it is a function with low levels of complex information. For example, in the random library used by Szostak, there were 4 sequences, out of 6×10^12 random sequences 80 AAs long, which showed weak binding to ATP, so that they could be selected by ATP binding columns.

    So, according to those data, we can conclude that the complexity of the function:

    “Any sequence 80 AAs long that can bind ATP, even with a very weak binding”

    is about 40 bits.

    Which is not a low complexity, but certainly not very high.

    So we can conclude that functions at that level of complexity can be found in realistic random repertoires. Those functions are the result of mere RV, like the stone on the beach.

    The simpler the function, the lower the complexity of specific information necessary to implement it.

    So, let’s go back to our “low-medium” complexity function: weakly binding ATP.

    The second question is: can further functional information be added to it by a process of selection?

    Let’s see. We have the space of selection procedures, S, and we have generated a binary partition in it (see my previous posts in this thread, especially #275), so that we can recognize two subsets:

    NS (very small)

    AS (very big)

    In the case of our function, weak ATP binding, can those selection procedures intervene, acting on the few natural sequences exhibiting that function in our random library?

    For NS, the answer is: no.

    There is no way that a weak capacity of binding ATP can give some reproductive advantage in any reasonable biological scenario. If someone does not agree, that someone is invited to explain why.

    So, NS is out of discussion in this case.

    But AS can certainly intervene: after all, AS can define its function as it likes, and select and expand it as it likes. So, if we use ATP binding columns, we can certainly select those natural sequences.

    But can the process of AS add functional information add new functional information to the sequences?

    Again, the answer is yes. After all, Szostak did exactly that, transforming the original weak binding into some strong ATP binding, even with some basic folding of the molecule.

    We cannot deny that a function defined as:

    “any sequence which can bind ATP with a strong binding, for example at least of such and such”

    must be more complex than the original function:

    “any sequence that can bind ATP at all”

    because for the first function the target space is certainly smaller.

    So, we can conclude that in this case AS can add some functional information to an original random low-level function.

    OK, the discussion is longer than I expected. I will continue in my next post.

  285. 285
    gpuccio says:

    To all interested:

    So, let’s continue our discussion of point c):

    c) How is it that NS can generate functional information, even if in small quantities? Can AS do more, and how much more?

    We have already seen that AS can select some specific function which is naturally present in some random repertoire, and in some way, through rounds of variation and repeated AS, add some functional information to that function. I will come back to that in a moment.

    But can NS do the same?

    Yes, it can. But it is strictly confined to functions that affect reproductive success, giving some advantage.

    Now, those cases are exceedingly rare. While low level functions in general (like weak ATP binding, or any other weak and trivial biochemical affinity) are moderately represented in random sequences (see the 40 bits of functional information for weak ATP binding), functions that can give some definite reproductive advantage are certainly, as a rule, much more complex, and cannot easily be found in a random library.

    The few clear examples that we have of NS in action, indeed, are about slightly modifying some existing functional structure to get an advantage in particular contexts (see simple antibiotic resistance) or retrieving an existing function which has been intentionally “damaged” and maintained at very low levels (see the rugged landscape experiment), or something like that.

    For example, penicillin resistance can be acquired thorugh mutations of PBPs, membrane-bound D,D-peptidases. Penicillins are substrate analogues, and they bind to the enzyme and inactivate it because they are similar to its natural substrate, peptidoglycan. So, the accumulation of one or more mutations that alter the enzyme’s binding site can confer antibiotic resistance, even if in principle they are desctructive mutations.

    Behe has clearly identified that mechanism as the “burning bridges to prevent the enemy from coming in” strategy.

    Of course, degrading an existing structure is extremely easy if compared to building a new structure. We can say that the function:

    “degrading an existing structure, so that antibiotics cannot any more target it”

    is a function with very low functional information. A lot of variations can degrade the existing strucutre, and the target space is very big.

    That’s exactly the reason why the almost powerless NS can be effective in these scenarios: the functional information to be found is very low.

    I quote here my final statement form my post #87, to the precious Corey Delvine:

    I will continue not to be amazed at the power of destructive random variation: whoever has destroyed a house of cards with a very slight movement knows that concept all too well.

    On the other hand, whoever has built a house of cards with a single, slight movement, is certainly a remarkable individual!

    So, our first conclusion is that:

    Both AS and NS can add some functional information to some existing scenario. However, the range of AS is certainly much wider. NS, instead, can only act on functions that confer a reproductive advantage, so its scope is extremely limited.

    So limited that the existing models are almost always about degrading an existing function or partially retrieving an existing damaged function.

    But there is another important factor that should be discussed: the protein space landscape. I will do that in next post.

  286. 286
    Origenes says:

    GPuccio @285

    GPuccio: Both AS and NS can add some functional information to some existing scenario.

    I have a problem with this. Perhaps, you can clarify what you mean by “adding information”?

    My line of thinking is thusly:
    An individual organism can only receive information by a mutation, never by selection (NS or AS). Selection only acts on what already exists, so in what sense does it add functional information?

    The idea is, if I understand it correctly, that selection (AS or NS) can add functional information to a population — not an individual organism. Selection can spread existing information throughout a population over time. Does this equate to “adding information to a population”?

    One thing is for sure, selection does not create any information.

    What selection does is making a population more homogenous by subtracting information. Selection is in fact elimination. After each successful round of selection there is less variation in a population, and in an important sense less information.

  287. 287
    gpuccio says:

    Origenes:

    Yes, what you say is right. Of course, it is always RV that adds information.

    What I meant is that the process of RV+NS can add functional information, in the contexts that I have specified.

    IOWs, the role of NS is to expand a selectable step so that it is no more confined to the initial individual, but spreads to the population. So, the probabilities of having a new “beneficial” mutation that can be added to the first are much higher.

    IOWs, if we are discussing bacteria, and the final object has a two AAs beneficial mutation, if we assume that the first mutation is selectable, what happens is more or less:

    a) we have a population of, say, 10^12 bacteria with the original allele

    b) a single mutation happens by RV in one individual

    c) it is beneficial, so in time x it is fixed (expanded to the whole population)

    d) a second single mutation happens in the same allele, and the final form with two beneficail mutations is reached.

    Now, the role of NS in that kind of scenario is important, because if no NS acted on the first mutation, we would have one individual (or at list its individual clone) where the second mutation should happen. The probabilities here are extremely low (p1*p2).

    But if the first mutation is expanded to 10^12 individuals, then the seconf mutation has an increase in probability of about 10^12, because now there are 10^12 bacteria with the first mutation which can, by RV, receive the second mutation and reach the target.

    So, while it is absolutely vtrue that it is always RV which generates the novelty, it is also true that NS, in this extreme scenario, has a fundamental role, because it provides more realistic probabilstic resources. That’s what I mean when I say that NS, if and when it can act, “lowers the probabilistic barriers”.

    Probably I should have said:

    “Both AS and NS, cooperating with RV, can contribute to add some functional information to some existing scenario.”

    Indeed, AS too cannot generate new information, in irself: it is always RV that generates new information in all scenarios based on variation + selection.

  288. 288
    Origenes says:

    GPuccio @287

    Gpuccio: … the role of NS is to expand a selectable step so that it is no more confined to the initial individual, but spreads to the population. So, the probabilities of having a new “beneficial” mutation that can be added to the first are much higher.

    That is true of course, but what if there is no second new beneficial mutation that can be added to the first? IOWs what if the first selectable step has no follow-up and leads to a dead end? Or what if the original sequence was only a few mutations away from a breakthrough evolutionary discovery, but was led astray by NS due to a rather trivial selectable step?

    Gpuccio: … So, while it is absolutely true that it is always RV which generates the novelty, it is also true that NS, in this extreme scenario, has a fundamental role, because it provides more realistic probabilistic resources. That’s what I mean when I say that NS, if and when it can act, “lowers the probabilistic barriers.”

    I am sorry, but I am not convinced. It seems to me that probabilistic barriers are only lowered if, by sheer dumb luck, the area ‘chosen’ by NS happens to contain, nicely lined up, further beneficial mutations. Counting on this is obviously a huge gamble. With a varied population you have a broad search, which is narrowed down by NS, but, again, that’s quite a gamble. It is comparable to searching for an Easter egg on an island. One can have several small groups searching all over the island or one can take a huge gamble by forming one single large group and focus the search on a specific part of the beach.
    What is the better strategy? Which method “lowers the probabilistic barriers”?

    Gpuccio: Probably I should have said: “Both AS and NS, cooperating with RV, can contribute to add some functional information to some existing scenario.”

    Sure it can. Anything is possible if you are incredibly lucky — if selectable steps just happen to be nicely lined up.

  289. 289
    Dionisio says:

    gpuccio,

    Please, would you mind to comment on this?
    Thanks.

    “…all mutational types have worked in concert with evolutionary forces to generate the current human brain…”

    https://link.springer.com/content/pdf/10.1186%2Fs12915-017-0409-z.pdf

    May the [evolutionary] force[s] be with you” [Star Wars] 🙂

    [emphasis added]

  290. 290
    Dionisio says:

    The politely dissenting interlocutors seem practically gone from this interesting technical discussion thread, but apparently gpuccio and his follow up commenters (Origenes, Mung,…) are “cooking” something tasteful here…

    This thread keeps attracting readers in relatively larger numbers:

    Popular Posts (Last 30 Days)

    What are the limits of Natural Selection? An interesting… (2,234) [since October 5]

    Violence is Inherent in Atheist Politics (2,016)

    Selective Horrid Doubt (1,319)

    Sweeping the Origin of Life Under the Rug (1,004)

    Has the missing matter of our universe finally been found? (845)

  291. 291
    gpuccio says:

    Origenes:

    Again, I think that we agree on everything, but probably I have not made clear enough the thread of my reasoning, which however is still incomplete.

    Now I have not the time, but later in the day I will try to explain why we essentially agree, and if possible to complete my reasoning with the part about functional protein space. 🙂

  292. 292
    Dionisio says:

    gpuccio,

    I’m looking forward to reading the comments on the functional protein space landscape, though I think you’ve touched that topic before.

    Thanks.

  293. 293
    Dionisio says:

    gpuccio,

    This was posted first in another discussion thread started by News, but I’ll post it here too, though perhaps it’s slightly off topic within this thread:

    In the abstract of this paper referenced @289:

    Linker, S.B., Marchetto, M.C., Narvaiza, I. et al. BMC Biol (2017) 15: 68. https://doi.org/10.1186/s12915-017-0409-z
    https://link.springer.com/content/pdf/10.1186%2Fs12915-017-0409-z.pdf

    we can read this:

    “…there have been multiple waves of LINE retrotransposition as well as the birth of new mobile elements such as the SINEs Alu and SVA…”

    As far as you’re aware of, is there any literature explaining how the birth of new mobile elements may have occurred?

    It would be interesting to review such a literature and learn more about that important topic.

    Thanks.

    [emphasis added]

  294. 294
    Dionisio says:

    @293 follow-up

    […] all forms of mutation work together with selection and drift to produce the ever-dynamic phenotype […]

    Linker, S.B., Marchetto, M.C., Narvaiza, I. et al. BMC Biol (2017) 15: 68. https://doi.org/10.1186/s12915-017-0409-z
    https://link.springer.com/content/pdf/10.1186%2Fs12915-017-0409-z.pdf

    Any comments on the above quoted statement?

  295. 295
    Origenes says:

    GPuccio @291

    Do complete your very important reasoning first and please ignore my distracting remarks.
    I will refrain from commenting until you have made your point on functional protein space and AS.

  296. 296
    EugeneS says:

    A very interesting thread indeed!

  297. 297
    gpuccio says:

    Origenes:

    You say:

    That is true of course, but what if there is no second new beneficial mutation that can be added to the first? IOWs what if the first selectable step has no follow-up and leads to a dead end? Or what if the original sequence was only a few mutations away from a breakthrough evolutionary discovery, but was led astray by NS due to a rather trivial selectable step?

    I agree with you. That’s exactly what happens in the general case.

    “Beneficial” mutations are indeed exceedingly rare.

    And almost all of them are isolated, and lead to nothing else, whether they are selected or not.

    We agree on that! 🙂

    I am sorry, but I am not convinced. It seems to me that probabilistic barriers are only lowered if, by sheer dumb luck, the area ‘chosen’ by NS happens to contain, nicely lined up, further beneficial mutations. Counting on this is obviously a huge gamble. With a varied population you have a broad search, which is narrowed down by NS, but, again, that’s quite a gamble.

    And again I agree: it’s quite a gamble! 🙂

    Then you can ask: why have I spent a lot of time discussing how NS (and AS) can in some cases add some functional information to a sequence? (see my posts #284, #285 and #287)

    There is a very good reason for that, IMO.

    I am arguing that:

    1) It is possible for NS to add some functional information to a sequence, in a few very specific cases, but:

    2) Those cases are extremely rare exceptions, with very specific features, and:

    3) If we understand well what are the feature that allow, in those exceptional cases, those limited “successes” of NS, we can easily demonstrate that:

    4) Because of those same features that allow the intervention of NS, those scenarios can never, never be steps to complex functional information.

    Which is the real reason that no one has even tried to take my challenge at #103.

    Because the few examples of microevolution where NS is somewhat successful are also the evidence that it cannot do anything more than that.

    The final conclusion is:

    5) NS, acting on RV, can never, never generate any complex functional information.

    Now, I will try to go on with my arguments, and to specify better the reasons for those conclusions.

  298. 298
    gpuccio says:

    EugeneS:

    “A very interesting thread indeed!”

    Thank you. 🙂

    I think we are touching some important issues here.

  299. 299
    Mung says:

    What does it mean to say that information is added or deleted. That is an interesting question and I commend gpuccio for trying to answer it.

    Perhaps information is not deleted but rather lost due to noise.

    As far as information being added. Say you have a source with an alphabet of four letters. When one of those four letters is selected for transmission, is that adding information? Is that what happens when DNA is copied or a point mutation takes place?

    This is probably a new thread on its own, lol.

    So what does natural selection have to do with that? Not much, as far as I can see. But here we get into one of the problems with selection. At what level does it operate? At the nucleotide level? At the level of the gene?

    If you have an allele with a given probability distribution and one of those alleles is selected, is that adding information?

    Thus we have no prestatable alphabet upon which to build any known information theory. Therefore, there is no sense in which information theory is applicable to biological evolutiion.

    – Stuart A. Kauffman

  300. 300
    Mung says:

    gpuccio:

    That’s what I mean when I say that NS, if and when it can act, “lowers the probabilistic barriers”.

    I disagree. 🙂

    The difference is not that natural selection lowers the probabilistic barriers, the difference is that the probability distribution is different.

    I guess I need to dig out some lessons on population genetics and make sure I’m not just making things up again, lol.

  301. 301
    gpuccio says:

    Mung:

    As far as information being added. Say you have a source with an alphabet of four letters. When one of those four letters is selected for transmission, is that adding information? Is that what happens when DNA is copied or a point mutation takes place?

    This is probably a new thread on its own, lol.

    So what does natural selection have to do with that? Not much, as far as I can see. But here we get into one of the problems with selection. At what level does it operate? At the nucleotide level? At the level of the gene?

    If you have an allele with a given probability distribution and one of those alleles is selected, is that adding information?

    Yes, that could be a new thread! 🙂

    For the moment, I will only say that I do not speake of information in general, but only of functional information in a sequence computed for one explicitly defined function.

    So, my meaning is well defined.

    “Adding functional information” therefore means that we get a sequence which can implement some function which is more complex (IOWs, that needs more specific bits to be implemented).

    So, if we go back to Szostak’s experiment, we have that he starts from sequences with a weak ATP binding, a function for which I have computed about 40 bits of functional information, from the data he gives us.

    Then, by round of nutation and artificial selection, he gets one protein with strong ATP binding.

    Now, according to my definitions, there can be no doubt that a protein with strong ATP binding has higher functional information, in relation to ATP binding, than a protein with weak ATP binding. Of course, weak ATP binding is certainlt less specific, and requires less specific bits to be implemented.

    So, we can say that the process of AS realized by Szostak in his lab has added some functional information, always related to ATP binding, to the original 40 bits of the random sequences selected from the random library, although I have not enough data to say how many bits were added (but it could be done, if we knew all the details of the procedures).

    That’s what I mean by “adding functional information”. It’s a very specific, explicit and objective meaning.

  302. 302
    gpuccio says:

    To all interested:

    OK, the starting point for this part of the discussion was:

    c) How is it that NS can generate functional information, even if in small quantities? Can AS do more, and how much more?

    Let’s answer the two issues separately.

    How is it that NS can generate functional information, even if in small quantities?

    The key word here is: “how?”

    We have already seen that new functions that can give some detectable reproductive advantage are as a rule too complex to be found in random libraries: IOWs, RV in itself cannot generate them.

    So, while we can find in a random library the function:

    “weak ATP binding”

    which is simple enough (about 40 bits), but certainly not naturally selectable, we cannot certainly find in random library a function like:

    a) Having a strong ATP binding that:

    b) Implies ATPase activity that:

    c) Releases biochemical energy and:

    d) Couples that energy to:

    e) Some complex biochemical machine that uses it to get the functional result.

    That kind of thing is probably naturally selectable (if the final result confers a reproductive advantage, but where can we find such a thing in a random library?

    The minimal functional complexity of such a machine is probably of hundreds of bits!

    Remember the results of Axe and of the rugged landscape experiment, which both mention starting libraries of the order of 10^70 to be able to get something really meaningful from a random sequence!

    So, our first conclusion is:

    1) We cannot find, as a rule, any new naturally selectable function by RV alone.

  303. 303
    gpuccio says:

    To all interested:

    So, we come to an important point:

    The cases of microevolution by RV and NS that we know of have a first common feature: NS does not act on some new function, but always on some small RV of some existing functional structure, usually already very complex

    As you can see, that is very different from what happens in the cases of AS we know of. For example, Szostak acts by AS on rnadom sequences from a random library already exhibiting a weak ATP binding.

    So, in the starting scenario, weak ATP binding is a “new” function, and it arises in the library by mere chance.

    Of course, as we have said, it is a rather simple and absolutely non naturally selectable function!

    So, let’s go back to NS.

    The simplest case could be penicillin resistance by mutations in PBPs.

    PBPs are bacterial membrane-bound D,D-peptidases that catalyze the transpeptidation reaction that cross-links the peptidoglycan of the bacterial cell wall. Penicillins act as substrate analogues and covalently bind to the PBP active site serine, thus inactivating PBPs.

    The simplest case of penicillin resistance is probably what can happen at the level of a PBP molecule. I quote from the paper:

    Penicillin-Binding Protein-Mediated Resistance in Pneumococci and Staphylococci

    https://academic.oup.com/jid/article-lookup/doi/10.1086/513854

    In the laboratory, very low-level resistance to penicillin can result from a point mutation leading to amino acid substitution within the transpeptidase domain of the PBP. Mutations described in the PBP 2x gene of penicillin-resistant mutants of S. pneumoniae R6 are instructive [6]. Mutations occurred within the penicillin-binding domain of the PBP 2x molecule (figure 1), often in regions near one of the three penicillin-binding motifs, SXXK (which contains the active site serine), S(Y)XN, and K(H)T(S)G. One point mutation resulted in a modest increase in MIC from 0.02 to 0.16 ?/mL (table 2). A second mutation doubled that MIC to 0.32. Additional mutations in PBP 2x were associated with higher levels of resistance, with MICs of 1.28 ?g/mL.

    This is a simplified scenario, and indeed things are probably much more complex.

    However, I mention it to show a realistic pathway where:

    a) One single point mutation can confer some resistance to penicillin, probably altering the affinity between the PBP anf the antibiotic.

    b) Other (few) point mutations can increase the resistance.

    So, in this scenario we could have:

    a) A first event which is compeltely random, and acts modifying slightly an existing functional structure (the PBP), and conferring indirectly an advantage (lower affinity to penicillin). One point mutation is a very simple event, in bacteria it can certainly happen rather frequently, and it is potentially selectable by NS under antibiotic treatment, because of the low resistance it confers.

    b) Once the basic function of resistance is there, it can be optimized by further simple mutations.

    However, the sum total of effective mutations in this scenario remains low (maybe 3 – 6).

    I will try to draw now the important features that allow NS to generate some functional information:

    1) As already said, the basic function arises always from rather simple modifications to existing, complex, functional structures (in this case, the PBP).

    2) The original simple modification arises from pure RV. Therefore, it must be really simple (in this case, a single point mutation).

    3) In most cases the original modification is more or less deleterious to the original function of the complex function structure. Indeed, a mutation to a PBP is very likely to be deleterious to the normal biochemical function of PBP, and very often it is.

    4) However, in cases of sever abnormal selection pressure (in this case, antibiotic treatment) the advantage deriving from resistance can prevail on the possible deleterious effect of the mutation.

    5) NS can therefore fix the original mutation.

    6) A few successive simple random events can optimize the resistance (or any “new” function derived from the slight modification of the existing complex function). In all cases we know well, however, the total number of mutations required for the optimization remain very low (3-6), and therefore the new functional information generated in this type of context remains always very low (6 AAs mean about 24 bits at most). No complex functional information is generated, but certainly a small amount of functional information is generated.

    7) As shown by our example, the functional space that goes from the original resistance (1 mutation) to the final optimized resistance (let’s say 6 mutations) is rather continuous: we can reasonably assume that, in this scenario, each single new mutation increases the resistance, and can be selected.

    As we will discuss, this is usually the case with very simple functions.

    That’s why this kind of resistance is rather frequent. However, as I have said, this is a simplified scenario. The full scenarios are more complex, and not always well understood. Here is a more recent paper that can giv some specific information:

    http://onlinelibrary.wiley.com.....095.x/full

    OK, in next post we will apply these concepts to another famous scenario of antibiotic resistance, chloroquine resistance in plasmodium falciparum.

  304. 304
    mike1962 says:

    Excellent OP and thread!

    I have particularly enjoyed the exchange between Gpuccio and Corey. Especially Corey @199 where he leads with his chin, with the ensuing aftermath of his destruction by Gpuccio, leading to Corey sputtering, literature bluffing, and up shifting his insults to Gpuccio. These types are so predictable.

    Thank you Gpuccio for all the time you put into this. There are many of us lurkers out here who are very appreciative of your good work.

  305. 305
    Dionisio says:

    gpuccio @303:

    The cases of microevolution by RV and NS that we know of have a first common feature: NS does not act on some new function, but always on some small RV of some existing functional structure, usually already very complex

    This is a very concise and clear description of a fundamental concept that should be kept in mind when we look at the evo-devo literature too. Thanks!

  306. 306
    Dionisio says:

    mike1962 @304

    “Excellent OP and thread!”

    Yes, as EugeneS wrote @296, this is a very interesting thread indeed!

    Regarding the politely dissenting interlocutor you mentioned, apparently that folk couldn’t withstand the heat of the discussion under the weight of gpuccio’s strong technical arguments, so the interlocutor ran quickly for the door. Poor thing. 🙂

  307. 307
    Eugene S says:

    GP

    Actually, speaking about answering challenges… My OP that I wrote back in April was also met with cold silence from the other side.

    Sigh.

    I wonder if the silence is intended to mortify the opponent by contempt? 😉 A powerful weapon indeed.

  308. 308
    gpuccio says:

    mike1962:

    Thank you! 🙂

    Your very competent opinion is always precious.

  309. 309
    gpuccio says:

    Eugene S:

    Maybe contempt is not such a bad thing. You know, the perfect joy story and so on…

    By the way, I will read your OP as soon as I have a couple of minutes. It seems really interesting! 🙂

  310. 310
    gpuccio says:

    Dionisio:

    Many thanks to you too, for your constant support!

    Well, Corey was indeed a good source of debate… 🙂

  311. 311
    gpuccio says:

    To all interested:

    Now let’s go to chloroquine resistance in plasmodium falciparum, another classic example of microevolution by RV and NS.

    Of course, Behe has discussed that in detail in TEOE. Many have criticized his concepts. A post by Moran about the issue has been quoted by Gordon Davisson at the beginning of this discussion. Here it is:

    http://sandwalk.blogspot.it/20.....ution.html

    For the following discussion I will refer directly to the paper by Summers et al. quoted by Moran, which is:

    http://www.pnas.org/content/111/17/E1759.full.pdf

    It is a very good and recent (2014) paper, which allows to have very good insights into the process of microevolution that is described in it.

    In brief, the premise is that chloroquine is concentrated in the digestive vacuole of the parasite, and there it exerts its toxicity.

    Now, chloroquine resistance (CR) is rare, much rarer than common antibiotic resistance (including resistance to other anti-malaria drugs).

    As we all know, Behe explains that simple fact assuming that CR requires two independent mutations, and not only one, to arise.

    The protein that is mutated in CR is the parasite protein PfCRT, a 424 AAs long protein in the membrane of the digestive vacuole, whose natural function is not known: it is probably some kind of transporter, a channel or a carrier.

    However, in its mutated form, it can transport chloroquine out of the vacuole, and that is the basis for CR.

    In the study, the authors examined 7 variants of mutated PfCRT that arose independently in different part of the world.

    Figure A shows the mutations in the 7 variants, in reference to the wildtype (HB3), which has no CR. The minimum number of mutations is 4 (Ph1), the maximum 8 (Dd2).

    The authors have tested about 60 variants of those molecules, artificially generated, to understand the role of each mutation in conferring CR. A very good analytical work!

    The results are summarized in Fig. 2 (A and B), in Table 1 and especially in Fig. 3 (which is the one quoted by Moran).

    They found that there are two different mutational routes by which CR is acquired, and they called them ET and TD.

    Both routes start with two mutations:

    a) The ET route starts with 75E and 76T. It corresponds to the artificial variant named D32 in Fig. 2A. It has low but significant CR (about 19% of the final resistance in the Dd2 variant).

    b) The TD route starts with 76T and 326D. It corresposnds to the artificial variant named E6 in Fig. 2B. Again, it has low but significant CR (comparable to that of D32).

    As you can see, mutation 76T is common to both starting routes. Indeed, it is common to all CR variants. It is absolutely needed for CR.

    However, mutation 76T alone does not confer any CR, as you can see in variant D38 (Fig. 2B).

    Mutation 75E alone gives no CR, as can be seen in variant D39 (Fig. 2A).

    Mutation 326D alone, without 76T, gives no CR, as can be seen in variant E1 (Fig. 2B).

    So, to make it brief, the simplest way to have CR at low but significant level is:

    the 76T mutation (which is always required)

    and, associated with it:

    either the 75E mutation (route ET)

    or the 326D mutation (route TD)

    So, a first important conclusion can be reached by these simple facts:

    Chloroquine resistance requires two independent mutations to arise, none of which can confer any resistance if it happens alone

    So, here we are in a different scenario from other kinds of antibibiotic resistance which can start with one single point mutation.

    Here, two independent mutations, each of which is at least neutral in regards to CR, must be generated by RV alone in the same individual parasite. Otherwise, no CR is present, and therefore NS cannot act.

    IOWs, this first step, which already confers almost 20% of the final maximum function, must happen by RV alone!

    That’s why this kind of resistance is much rarer than other forms of antibiotic resistance, which arise often because their starting event is a single mutation.

    The following intervention of NS (which I will discuss in next post) has nothing to do with that difference: the difference is in the complexity of the starting event.

    IOWs, for chloroquine resistance we have to “wait for two mutations”, while in other forms of antibiotic resistance we only have to “wait for one mutation”, for the function to appear.

    The following process which optimizes the starting function by adding a few more mutations with the help of NS is comparable in both kinds of resistance, but the starting event is much more unlikely in chloroquine resistance!

    IOWs, Behe was completely right!

    More in next post.

  312. 312
    gpuccio says:

    EugeneS:

    I see that in your OP you say the same things that I am saying here, even if in a different context!

    Good. Really good! 🙂

    Natural selection is nowhere near intelligent selection in efficiency. Natural selection is most efficient and sensitive when selective pressure is acting with respect to a very limited number of traits (optimization criteria).

  313. 313
    EugeneS says:

    GP #312

    Great! Thanks for your comment.

    Exactly, RV+NS is really able to produce information noise. Anything nontrivial requires insight, foresight, planning, decision making, top-down strategy.

  314. 314
    Origenes says:

    The Case Against a Darwinian Origin of Protein Folds‘ by Axe, is required reading in order to understand what GPuccio is talking about.

  315. 315
    gpuccio says:

    Origenes:

    Yes, that’s a very good reference! Thank you. 🙂

  316. 316
    gpuccio says:

    To all interested:

    So, let’s go on with our chloroquine resistance scenario.

    Whay happens when the first “step”, which already gives a significant resistance (almost 20% of the total) appears by mere RV? IOWs after we have got after the “waiting for two mutations” phase?

    Of course, we can expect that the new mutated protein confers a well detectable reproductive advantage under the sever pressure of chloroquine administration in the patien. After all, chloroquine has the power to kill amost all other parasites, and thereofre the individual which got the resistance, and its descendants, are greatly favoured in that context.

    So, we can expect that the selection coefficient of the new trait, in that specific environment, will be very high. It is very likely that the new trait will be fixed in that population, and probably it will be fixed rather quickly, if the administration of chloroquine remains a rather frequent environmental factor.

    So, our next step is: NS can fix the new trait derived from two random mutations.

    At that point, new single mutations can happen, with a rather reasonable probability.

    It is important to understand that at this pint, if we had to wait for two mutations again, the evolutionary times, alredy rather long, would be doubled juist to reach a new step.

    But, if we have to wait only for one mutation, the new waiting time will be reasonably short.

    Is that the case in chloroquine resistance?

    It seems so.

    Let’s look now at Fig. 3 of the Summers paper, the same quoted and “interpreted” by Moran on his blog. For once, we will not use Moran’s comments, and will try to understand what the paper says with our limited intelligence. 🙂

    I will try to give a brief summary, and then the details.

    It is possible to get the variants with high CR thorugh rather simple pathways.

    a) All pathways require the initial “lucky” random step of two specific mutations.

    b) From that starting point, all pathways proceed by single selectable mutations, in the number of 2 or 3.

    c) so, 4 total mutations, or in some cases 5, are necessary to get a high level of CR.

    Let’s see more in detail.

    Let’s start with the pathway on the right (Fig. 3 B), which leads to the variant Ecu1110. This is a variant present in nature, which has a good, but not maximal, CR. In Figure 2 B, we can see that it is about 36% of Dd2.

    This form can be derived by the TD route, which starts with:

    2 random mutations: 76T and 326D (variant E6, CR about 15% of Dd2)

    by adding two successive single mutations:

    + 356L (variant E2, CR about 29% of Dd2)

    + 220S (final target, Ecu1110, CR about 36% of Dd2).

    As we can see, each of those two individual mutations is naturally selectable, because each optimizes the existing function.

    So, the whole pathway to Ecu1110 can be summarized as follows:

    1) Appearance in one individual parasite with the wildtype sequence (HB3) of the E6 variant with 2 mutations (76T and 326D):

    waiting time: “waiting for two mutations”

    explanation: RV

    2) Fixation of E6:

    waiting time: probably short

    explanation: NS

    3) Appearance in one individual parasite with the E6 sequence of the E2 variant with 1 mutation (356L):

    waiting time: “waiting for one mutation”

    explanation: RV

    4) Fixation of E2:

    waiting time: probably short

    explanation: NS

    5) Appearance in one individual parasite with the E2 sequence of the final target Ecu1110 with 1 mutation (220S):

    waiting time: “waiting for one mutation”

    explanation: RV

    6) Fixation of Ecu1110:

    waiting time: probably short

    explanation: NS

    That’s all: 4 mutations in all, 2 of them beneficial in couple, 2 of them beneficial. 3 fixations by NS.

    Critical time: the initial “waiting for two mutations” time.

    More in next post.

  317. 317
    gpuccio says:

    To all interested:

    Now that we know the procedure, let’s see briefly the pathway to D17. This is a variant which is not present in nature, but it has the same level of CR as the “natural” Dd2.

    The pathway is the same as the one we have already analyzed, it requires 4 total mutations, but it starts from the ET route.

    So we have, starting from the wildtype:

    a) 2 random mutations (75E + 76T), in couple: D32, about 20% of Dd2.

    b) + 1 random mutation (271E): D27, 55% of Dd2.

    c) + 1 random mutattion (220S): final target, D17, 100″ of Dd2.

    Again, 3 interventions of NS.

    More in next post.

  318. 318
    gpuccio says:

    To all interested:

    Very briefly again:

    The path to K1, the natural variant with the highest CR (120% od Dd2), is similar, but it requires one more single RV + NS. It starts from the ET route (75E + 76T), and requires 3 more successive beneficial mutations: 220S, 74I, 326S. % mutations in all, 4 interventions of NS.

    The K1 variant also presents 2 more mutations (371I and 271E), which are neutral to the CR function, and are therefore merely accidental.

    Finally, the natural Dd2 variant can be reached by some more complex pathways, but in the end it also includes 5 necessary mutations in the final form (the usual 2 of the ET route, + 3 more).

    More in next post.

  319. 319
    gpuccio says:

    To all interested:

    Now I would like to make some comments which can be of some importance.

    The first is, IMO, very important.

    The pathways that I have discussed in the last 3 posts are:

    explicit pathways which explain how a functional target can be reached from some starting sequence by RV and NS. As they are explicit, we can analyze the logic, and compute the probabilities, and decide if they are realistic and good explanations of what we observe.

    As for me, I would definitely say that they are.

    That conclusion is important, because it demonstrates that it is possible to show explicitly that some neo-darwinian pathway really exists. When it exists.

    The whole point of this long discourse is to say that: if a darwinian pathway exists, we can give a detailed conceptual description of the pathway itself, and find empirical support for it.

    In these well known examples of microevolution, like chloroquine resistance, darwinian pathways definitely exist. Therefore, they can be made explicit. There is no need of just so arguments, of imagination, of blind faith. In this well known examples, the neo-darwinian algorithm is a good scientific explanation for what we observe.

    On the other hand, no explicit pathways have ever been even vaguely suggested for any protein sequence that exhibits complex functional information. None at all.

    And the reason is very simple: because those pathways do not exist.

    So, the right question at this point is:

    Why do darwinian pathways exist in those few well known cases of microevolution, while no such thing exists for proteins exhibiting complex functional information?

    IOWs, why was I so sure that nobody would have answered my two simple questions in my challenge at #103?

    Many reasons have already been presented. But I will try to sum up and deepen that issue in next post.

  320. 320
    gpuccio says:

    To all interested:

    So, let’s see which are the specific features present in practically all the best knwon examples of generation of some functional information by the RV + NS algorithm, practically the only real evidence for neo-darwinism as an explanatory theory:

    1) The starting point of the process is always an already existing, complex and functional structure.

    2) The first step is always effected by RV only: it is alwatys a small variation of the pre-existing structure. That variation is usually of one single mutation (4 bits at most), but we have good examples of two mutations, (8 bits most), like in chloroquine resistance. I am not aware of initial random steps of 3 mutations, but they could in principle exist. I doubt that 4 random mutations steps are feasible.

    However, the point is not so much where exactly the edge is. The point is that the probability of a first random step decreases exponentially with each new mutation necessary to reach the basic selectable function. We know that in chloroquine reistance the frequency of the two mutations event is dramatically lower than the frequency of a single mutation event in other antibiotic resistance scenarios. The paper “waiting for two mutations” gives us a good idea of how difficult a two mutations event is in reality.

    So, let’s calll this concept the edge of random evolution. I believe that Behe is correct in proposing two mutations: that is probably the real edge in the general case. But I can grant 3 mutations, in exceptional cases, and even 4 mutations as an extreme value, just to be generous.

    So, let’s say that the edge of random evolution is 2 – 4 mutations (at most, 8 – 16 bits).

    That basic starting function can now be naturally selected and fixed, with a probability which depends on many factors (I will be back on that).

    3) If there is a continuous functional space around the original basic starting function, a few new mutations can accumulate, optimizing the original function.

    What is a continuous functional space?

    It’s simple. In this case, it just means that there are, among all the possible one aminoacid substitutions, one or more that improve the original basic function.

    In the case of CR, we have seen that such a thing is true. Indeed, there is more than one possible selectable pathway.

    It is important now that each successive step be simple (one mutation only), to avoid that new huge improbabilities accumulate for the process. Indeed, in the case of CR, all the successive selectable steps are of 1 mutation.

    Each successive step must be naturally selectable, IOWs it must improve the original function. So, a ladder of naturally selectable and simple (1 mutation) steps can be available to reach the final target.

    So, how many mutations can be added by NS?

    In the case of CR, we have observed only pathways that add 2 or 3 mutations to the original 2, for a total of 4 or 5.

    In the case of penilcillin resistance in pneumococcus, I quote from the recent paper I already referenced:

    The individual reversion of the 41 substitutions of a transpeptidase domain of PBP2x from a particularly resistant strain uncovered the role of four mutations, I371T, R384G, M400T and N605T, in addition to those in positions 338 and 339 (Carapito et al., 2006) (Fig. 8). Reversion of the six substitutions nearly restored a normal and rapid acylation rate. Introduction of five of the substitutions into PBP2x from a susceptible strain diminished the reactivity with ?-lactams, almost to the level of the original PBP2x from the resistant strain. These effects measured in vitro were mirrored by the expected phenotypic consequences in vivo (Carapito et al., 2006). With a different PBP2x, a similar experiment where the effect of individual reversions was directly monitored in vivo also identified positions 371 and 384 as important (Smith & Klugman, 2005).

    Here, too, the number of useful total mutations that can optimize the resistance is in the same range: 2 – 6.

    So, we can define that as the edge of simple optimization : in well known cases, let’s say it is 6 total mutations, IOWs adding, with the help of NS, 4-5 single mutations to the original 1-2 mutations random event.

    6 total mutations is a functional information of, at most, 24 bits.

    A very simple concept should be clear: in all these well known cases, the intervention of NS acts only to otpimize the existing new function which was created in the first step as a slight random modification of the original functional structure: in the case of antibiotic resistance, the modification of an existing function which indirectly confers some resistance ot the antibiotic.

    That brings us to the next point:

    4) All these well known cases can be defined as cases of loss of function mutations which indirectly confer an advantage in a specific selective environment.

    Let’s call this kind of mutations: beneficial loss-of-function mutations.

    The term is not mine. I quote from this recent and interesting paper:

    “The spectrum of adaptive mutations in experimental evolution”

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4268012/

    The availability of beneficial loss-of-function mutations and the large target size for these events ensures that these mutations will come to dominate experimental evolution over short time scales.

    Emphasis mine.

    The point is very simple: loss of function mutations are very simple, even those which are in a sense “beneficial”. That’s why their “target size” is large. That’s why they are in the range of RV.

    We have seen that RV can give us, in most cases, only 1 or 2 mutations complexity. We have conceded 4, for the sake of discussion.

    That is enough to generate a loss of function in an existing structure, and in some cases an indirect beneficial effect through that loss of function.

    But it is definitely not enough to generate a new function from scratch!

    Not enough at all!

    And believe me, it’s not a question of the “short time scale”. Those guys have their math wrong.

    The important point is:

    Time scales improve the probabilistic resources linearly

    while:

    Adding new necessary mutations to the first step basic function increases the probabilistic barriers exponentially!

    There is no hope at all for RV, in whatever time scales, if we have to reach thresholds of, say, 35 AAs (140 bits). But even 10 AAs (40 bits) would be inconceivable.

    Remember, Szostak found a 40 bit function in a rather vast random library: weak ATP binding. But that was not a naturally selectable function, not at all. Only artificial selection, set up by intelligent researchers, could detect and select that kind of function.

    Who has ever found a 40 bit naturally selectable function in a random library?

    So, our important conclusion here is that:

    Only very simple functions can be the first step of a pathway, because the first step of a pathway can rely only on RV, and the power of RV in terms of generating functional information is really minimal.

    That’s why the only real examples that we can find are of the type “beneficial loss-of-function mutations”, because that kind of basic functions are the simplest type.

    More in next post.

  321. 321
    Dionisio says:

    gpuccio @319:

    “…no explicit pathways have ever been even vaguely suggested for any protein sequence that exhibits complex functional information. None at all.

    And the reason is very simple: because those pathways do not exist.”

  322. 322
    gpuccio says:

    To all interested:

    Still two important points:

    5) The function that arises as the first random step, and is then optimized, must be directly and strongly related to survival and/or reproduction.

    In all well known cases of microevolution, the function is directly related to survival: antibiotic resistance, the rugged landscape experiment.

    That brings us to the final point:

    6) The environmental pressure must be extreme: antibiotic resistance arises so easily because the antibiotic that pervades the system is a direct cause of extinction for all non resistant forms of the population.

    The advantages of points 5) and 6) are rather obvious: the selection coefficient of the new trait is extremely high.

    IOWs, the new trait can be selected and fixed with great efficiency, and the time to fixation is relatively short.

    IOWs, this is the neo-darwinian algorithm at its best: huge populations, high reproduction rate, extreme emvironmental pressure, very simple functions that arise from 1 or 2 mutational events and can easily be optimized by single mutation events, by a definite ladder of increasing function.

    Piece of cake, indeed!

    And yet, CR is not so easy at all, as we have seen. It already imposes severe restrictions to the algorithm, and the simple bottleneck of waiting for two mutations makes it a rather rare event, even in those favorable settings.

    Now, a few words about optimization by a continuous functional landscape.

    We have seen that, in well known cases of microevolution, the optimization proceeds for a few steps, and then stops.

    There are, I believe, two strong reasons for that:

    a) The continuous landscape is limited, it is just a small neighbourhood of the simple starting functional island. Moreover, such a continuous neighborhood is probably more easily found in very simple functions.

    b) The optimization can only go that far. Starting from that functional island, after a few optimizing mutations, any new change can only be neutral or deleterious.

    So, we can reasonably believe that, in the case of CR, the few natural forms of resistant molecules represent the best that can be done.

    It’s not a case, I believe, that those rare cases of spontaneous resistance, arising separately from 2 different starting islands, have similar levels of functionality.

    If you look at Fig. 2A of the Summers paper, you can see that all the “natural” forms deriving form the ET route (Dd2, 783, K1, GB4, China e) have resistance rather comparable to that of Dd2.

    The natural forms arising from the TD route (Ecu, 7G8, Ph1, Ph2) have comparable values of resistance too, although lower than the values in the ET route.

    What does that mean?

    It means that those forms have probably reached the best optimization possible, starting form their respective initial island.

    And it means that the ET initial island has better possibilities of optimization than the TD initial island.

    So, we can learn some important concepts from these data:

    a) The routes of optimization of these simple starting functions seem to be rather short: a roof is quickly reached, and nothing better can be done.

    b) The initial island is imporant: it conditions that roof of optimization that can be reached.

    c) Even for a very simple function like CR, we have two different initial islands.

    d) The functional space between those two islands is not continuous. You have to start either from the ET route, or from the TD route, and follow the respective pathways. You cannot mix the two routes. I quote from the paper:

    These two mutational routes are referred to henceforth as “ET” (referring to 75E and 76T) and “TD” (referring to 76T and 326D). Somewhat surprisingly, the combination of N75E and N326D resulted in a decrease, rather than an increase, in CQ transport activity; the addition of N75E to PfCRTEcu1110 (C9) or S326D to PfCRTDd2 (C14) significantly reduced CQ uptake.

    We have here the simplest form of rugged landscape: different functional islands, minimally isolated, that can implement the same function with different levels of optimization.

    Well, in next post I will discuss some partially wrong (IMO) interpretations of these data by Gordon Davisson and Larry Moran.

  323. 323
    Origenes says:

    GPuccio provides us with some excellent arguments as to why there is no continuous functional landscape, and the paper by Douglas Axe (see #314) contains some more, but let’s suppose, for the sake of argument, that there is a continuous landscape. How would we explain that fact?

    Stuart Kaufmann: If mutation, recombination, and selection only work well on certain kinds of fitness landscapes, yet most organisms are sexual, and hence use recombination, and all organisms use mutation as a search mechanism, where did these well-wrought fitness landscapes come from, such that evolution manages to produce the fancy stuff around us?

    If the fitness landscape were such that it steers blind processes to discoveries far beyond our comprehension, what would explain that scenario?

    Dembski: The fitness landscape supplies the evolutionary process with information. Only finely tuned fitness landscapes that are sufficiently smooth, don’t isolate local optima, and, above all, reward ever-increasing complexity in biological structure and function are suitable for driving a full-fledged evolutionary process. So where do such fitness landscapes come from? …

    Okay, so the environment supplies the information needed to drive biological evolution. But where did the environment get that information? From itself? The problem with such an answer is this: conservation of information entails that, without added information, biology’s information problem remains constant (breaks even) or intensifies (gets worse) the further back in time we trace it.

    The whole magic of evolution is that it’s supposed to explain subsequent complexity in terms of prior simplicity, but conservation of information says that there never was a prior state of primordial simplicity — the information, absent external input, had to be there from the start.
    [source]

  324. 324
    Mung says:

    Origenes;

    …where did these well-wrought fitness landscapes come from, such that evolution manages to produce the fancy stuff around us?

    Can you tell me where this quote comes from?

    This same question applies to the work by Andreas Wagner.

    If the search space is constructed in a miraculous manner, how does that possibly exclude the miraculous?

  325. 325
    Origenes says:

    Mung @324

    The quote is from Kauffman — Investigations p.9.

    Mung: This same question applies to the work by Andreas Wagner.

    If the search space is constructed in a miraculous manner, how does that possibly exclude the miraculous?

    Indeed. It is an attempt to relocate the miraculous (the source of information) where it is less conspicuous: the environment, or in the case of Wagner a miraculous hyperdimensional cube search space.

  326. 326
    Dionisio says:

    Perhaps this is a biology case that answers gpuccio’s challenge @103 better than Mung’s sound-sensitive spot example?
    Here’s a set of three consecutive video presentations of proteins that apparently resulted from RV+NS producing new complex functional specified information (at least the presenter seems to imply that?) :
    1. https://www.youtube.com/embed/9RUHJhskW00
    2. https://www.youtube.com/embed/lVwKiWSu8XE
    3. https://www.youtube.com/embed/FRtqfpO8THU
    It seems like this might persuade me to switch sides in this debate?
    🙂

  327. 327
    gpuccio says:

    Dionisio:

    Of course, dynein! How could I not think of it!

    And yet, it is a very good candidate for one of my future OPs.

    It is so obvious that dynein arose from RV+NS. Perhaps we should really switch sides! 🙂

    (For all neo darwinists who may be devoid of sense of humour, we are just joking! 🙂 )

  328. 328
    gpuccio says:

    Origenes and Mung:

    Yes, that would be a super-miracle, indeed.

    But you know, when miracles happen, they become visible. They become facts.

    The simple problem with the idea of a miraculous protein space is that it’s nowhere to be seen, while resurrected Lazarus could certainly be seen!

    So, if the functional protein space is so miracuolus, how is it that there was no progression from the mutated PfCRT with good affinity for chloroquine to, say, some enzyme that can degrade chloroquine?

    The scenario is favourable, after all. We have this complex functional protein, a 424 AAs long proteins which does something that we still don’t understand well, and which has already acquired, by a few mutations, a good affinity for a new substrate, chloroquine.

    The protein already has some good folding: after all, it is a functional protein.

    So, there is no reason why it shoud not reach some distant functional island that confers the ability to degrade chloroquine, instead of simply transporting it out of the vacuole.

    What’s the problem? The protein functional space is so miraculous, after all! What’s one more little more miracle? Why does that kind of thing simply not happen?

  329. 329
    gpuccio says:

    To all interested:

    OK, now let’s go to Gordon Davisson and Larry Moran.

    In my OP, I quote Gordon Davisson as saying:

    This is simply wrong. Take the evolution of atovaquone resistance in P. falciparum (the malaria parasite). Unless I’m completely misreading the diagram Larry Moran gives in http://sandwalk.blogspot.com/2…..ution.html, one of the resistant variants (labelled “K1”) required 7 mutations in a fairly specific sequence, and at most 4 of them were beneficial. In order for this variant to evolve (which it did), it had to pass at least 3 steps unassisted by selection (which you claim here is impossible) and all 4 beneficial mutations had to overcome genetic drift.

    The discussion is, of course, about this page by Larry Moran:

    http://sandwalk.blogspot.it/20.....ution.html

    where he makes comments about the Summers paper that I have discussed here in detail. Most of the discussion is about Fig. 3 (A and B) from that paper.

    Now, you know how much I appreciate Gordon Davisson, but I must say that he makes a few errors here. I will try to clarify:

    1) The discussion is about chloroquine resistance, not atovaquone resistance. That’s only a minor mistake, of no importance, and I mention it only for the sake of clarity.

    2) The real problem is when he says:

    “In order for this variant to evolve (which it did), it had to pass at least 3 steps unassisted by selection (which you claim here is impossible)”

    He is speaking of the K1 variant.

    Now, it is absolutely not true that the K1 variant “had to pass at least 3 steps unassisted”. He is referring, of course, to the 3 neutral mutations that are found in the K1 variant if compared to the wildtype. Which, as we will see, are really only two, indeed one and a half.

    Let’s see: there are, indeed, 7 mutations.

    5 of them are beneficial:

    75E + 76T which, as we know, are beneficial only if in couple, and appear by RV

    220S, 74I, 326S which are added by RV + NS

    If you follow one of the possible pathways for K1 in Fig. 3 A, for example:

    HB3 – D39 – D32 – D30 – D20 – D10 – GB4 – K1

    we can see that the other two mutations present in K1:

    371I and 271E

    are neutral in that pathway.

    However, 371I is beneficial in another pathway, so we can consider it potentially beneficial in the right context.

    But 271E is always neutral.

    Now, my point is the following:

    If some mutation that we find in a final functional target is completely neutral to the function, there is absolutely no reason to state that the protein “had to pass that step unassisted” in its evolution. The simple truth is that that particular step is a mere accident.

    Indeed, the 271E mutation is present only in three natural variants, K1, Dd2 and GB4, which are of course related (they come from the same pathway), and in none of them it seems to contribute to the function.

    There is also no reason to believe that the neutral mutation had to be fixed by genetic drift. That is absolutely unnecessary. A neutral mutation that happpens to be found in an evolutionary pathway was probably simply “hitchhiked”.

    I quote from the paper:

    The spectrum of adaptive mutations in experimental evolution

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2271163/

    Whole genome sequencing has the potential to reveal all mutations in an evolved genome. Among these are the beneficial mutations, but also neutral or deleterious mutations that happen to arise in the same background and hitchhike to high frequency. This is particularly an issue in asexual populations where the entire genome is one linkage group

    IOWs, if some neutral, or even deleterious, mutation arises in the same individual where a beneficial mutation arises, it can be hitchiked by the NS that acts on the beneficail mutation, and therefore fixed only because it is linked to the beneficial mutation. Mere accident. It can happen or not happen, nothing changes. Neutral genetic drift has no role here.

    So, Gordon Davisson gives unnecessary relevance to the presence of neutral mutations in natural variants: if they do not contribute to the function, they mean nothing.

    But let’s go to Gordon Davisson’s last “error”:

    3) He believes too much in what Larry Moran says! 🙂

    But that brings us to the question of Larry Moran’s statements, which will be the subject of next post.

  330. 330
    gpuccio says:

    To all interested:

    Now, the important part:

    Where is Larry Moran wrong?

    Moran makes some precise statements about the Summers paper, and about Behe’s ideas.

    While I am ready to admit that he is in general respectful enough of Behe as a fellow biochemist, that does not prevent him from criticizing explicitly his ideas.

    I will refer here to Moran’s page about Summer’s paper, already quoted here many times:

    http://sandwalk.blogspot.it/20.....ution.html

    but also to a previous page of Moran about Behe’s ideas:

    http://sandwalk.blogspot.it/20.....ution.html

    In his more recent page, Moran summarizes Behe’s ideas rather vaguely:

    Behe uses the example of drug resistance in Plasmodium falciparum (malaria parasite). Resistance to atovaquone occurs quite often so that’s probably due to a single mutation. Resistance to chloroquine, on the other hand, is rare so it’s probably due to multiple mutations in the relevant gene (PfCRT, a gene that encodes a transporter protein)

    Indeed, in his older page, Moran had described Behe’s argument with greater precision and accuracy:

    Behe points out that it is sometimes very difficult for the malaria-causing parasite, Plasmodium falciparum, to develop resistance to some drugs used to treat malaria. That’s because the resistance gene has to acquire two specific mutations in order to become resistant. A single mutation does not confer resistance and, in many cases, the single mutation is actually detrimental. P. falciparum can become resistant because the population of these single-cell organisms is huge and they reproduce rapidly. Thus, even though the probability of a double mutation is low it will still happen.

    If the probability of a single mutation is about 10^10 per generation then the probability of a double mutation is 10^20. He refers to this kind of double mutation as CCC, for “chloroquine-complexity cluster,” named after mutation to chloroquine resistance in P. falciparum.1 Behe’s calculation is correct. If two simultaneous are required then the probability will, indeed, be close to 1 in 10^20.

    The emphasis on “simultaneous” is mine. I will soon explain the reason for that.

    Let’s go again to the more recent page. Moran goes on:

    The interesting part of his book was the correct claim that there was an edge of evolution and the incorrect claim that you can’t get chloroquine resistance by a stepwise, sequential route.

    Well, here is a very obvious error:

    Behe’s argument is not that “you can’t get chloroquine resistance by a stepwise, sequential route”. Not at all. It was, rather that chloroquine resistance required two independent mutations, each of them not selectable, and that therefore its probability was the product of the two individual mutation probabilities. As Moran himself had correctly described in his older page:

    “That’s because the resistance gene has to acquire two specific mutations in order to become resistant. A single mutation does not confer resistance and, in many cases, the single mutation is actually detrimental.”

    Again, the emphasis on “detrimental” is mine, and I will explain the reason for that soon.

    Now, I believe that Moran here is equivocating, more or less intentionally, on some important points. Let’s try to make them clear:

    1) As well described in Summer’s paper, we can say that the pathway to chloroquine resistance is made of two parts:

    a) a first step, which requires two neutral mutations

    b) successive steps, which are stepwise selectable (IOWs each single mutation confers an optimization of the existing resistance).

    That is what we know now, but Behe did not know all those details at the time he wrote his book.

    However, his argument, as recognized by Moran himself, was that chloroquine resistance required two independent mutations, each of them not individually beneficial, and therefore not selectable. That is the reason that explains why it is much rarer: in Behe’s (and Moran’s) words, if the probability of one mutation is 1:10^10, then the probability of two independent mutations is 1:10^20.

    So, Behe was completely right, and the Summers paper completely confirms his prediction.

    2) But then, why is Moran so critical of Behe’s statements?

    The simple truth is that he, more or less intentionally, misrepresents them. And then he criticizes his own misrepresentation.

    The misrepresentation is based on common misunderstandings of both probability theory and evolution theory.

    And it is realized thorugh the introduction of two inappropriate and misleading words:

    a) “simultaneously”

    b) “detrimental”

    a) Why does Moran say that:

    “If two simultaneous are required then the probability will, indeed, be close to 1 in 10^20.”?

    Where does that idea that the two mutations must be “simultaneous” come from?

    It is not true that the two mutations must be simultaneous, in the sense of occurring at the same time, or in the same individual.

    The simple requirement is that the two mutations must, at some time, be present “at the same time” in some individual in the population.

    The difference is huge.

    Let’s say that there is an individual in the population, le’s call it “a”, where one of the two mutations happen. The probability of that event is 1:10^10.

    There is no need that the other mutation should happen in the same individual, at the same time (IOWs, that it be “simultaneous”).

    What is needed is that, while the first mutation is passed on by “a” to its direct descendants, let’s call it “the a clone”, at some time the second mutation happen in one individual of that clone. A lot of time can pass before the second event. There is no need for simultaneity.

    What is needed is that, at some time, we have at least one individual with both mutations. Then, and only then, the new trait becomes selectable, because each of the two mutations in itself is neutral.

    This is exactly the scenario described by Summers for chloroquine resistance.

    And in this scenario, the probabilities multiply: if the probability of the first event is 1:10^10, the probability of having the two independent events in the same individual clone, whatever the time needed for that, are of about 1:10^20.

    No need for simultaneity. That is a false concept introduced by Moran, nothing else.

    The only probabilistic requirement is that the two events must be independent. That means that the first event does not improve the probability of the second one. And that is true, because the first event is not selectable, therefore it does not influence the probabilities of the second event, because it has no effect on the probabilistic resources.

    b) Why does Moran say that:

    “That’s because the resistance gene has to acquire two specific mutations in order to become resistant. A single mutation does not confer resistance and, in many cases, the single mutation is actually detrimental.”

    Why the attention to the possibility that one or even both mutations can be “detrimental”?

    That is certainly possible, but it is not part of the main argument. The main argument is that the two mutations are neutral.

    If the two mutations are neutral, then the probability of both occurring is p*p, 1:10^20 in our example.

    If even one of the two events is detrimental, the probabilities will be much lower.

    But that’s not Behe’s argument. Behe’s argument is that the probability of two neutral mutations is the product of each individual probability.

    Now, in his older page, Moran does not explicitly state his argument against Behe’s conclusions. He just opens a debate in the comments.

    But, if we look at his personal comment labeled:

    “Wednesday, October 06, 2010 4:11:00 PM”

    we can understand what he is proposing:

    Steve LaBonne says,

    Allowing multiple rolls of only ONE of the dice will still make Behe’s number way off. This doesn’t help him much unless BOTH mutations are SUFFICIENTLY detrimental to be subject to strong purifying selection. Only that constraint would be sufficient to require that they be (nearly) simultaneous.

    I think you’re close to understanding the main problem with Behe’s argument.

    He assumes that deleterious mutations will always be rapidly eliminated from the population. That’s consistent with the common understanding of evolution so it appears to set up an insoluble problem.

    However, you and I (and many others) know that evolution doesn’t work that way. There’s a lot of sloppiness and accident so it’s quite possible for inefficient proteins to hang around for a long time. You could get the same effect by gene duplication and messing with the spare copy.

    Again, the emphasis is mine.

    As you can see, he is criticizing Behe for “assuming that deleterious mutations will always be rapidly eliminated from the population.”

    It’s the “deleterious” misrepresentation.

    But Behe is not assuming that. Not at all. His computation of the probabilities refers to two neutral mutations, not to two deleterious mutations.

    OK, no more time now. I will add some more comments about Moran in next post, as soon as I can.

  331. 331
    Dionisio says:

    gpuccio @329:

    Please, don’t be so strict, those two names spell almost identically:

    chloroquine (11 letters)
    atovaquone (10 letters)

    6 common letters: 2 ‘o’ + 1 ‘q’ + 1 ‘u’ + 1 ‘n’ + 1 ‘e’
    only 5 different letters

    🙂

  332. 332
    gpuccio says:

    Dionisio:

    Yes, but did the changes happen simultaneously? 🙂

  333. 333
    gpuccio says:

    To all interested:

    Let’s finish this analysis of Larry Moran’s statements.

    In his more recent page, he says:

    The interesting part of his book was the correct claim that there was an edge of evolution and the incorrect claim that you can’t get chloroquine resistance by a stepwise, sequential route.

    Why incorrect? It is perfectly correct!

    Of course, as we have already said, and as it is clearly showed in the Summers paper, the pathway to chloroquine resistance is made up of two parts:

    a) a first step, which requires two neutral mutations that must happen in the same individual or clone, and that must be present at the same time for NS to act. Each mutation is independent from the other one, and none of the two mutations can be selected if isolated. IOWs, both mutations are neutral, is isolated, in regard to chloroquine resistance.

    b) successive steps, which are stepwise selectable (IOWs each single mutation confers an optimization of the existing resistance).

    Now, while the steps in b) are certainly stepwise selectable, chloroquine resistance cannot happen at all if a) does not take place as a first step.

    The two mutations that make up that a) step, therefore, cannot be obtained in a “stepwise, sequential route”. There is no sequenctial route to them least of all stepwise, least of all sequential.

    They must just happen, independently one from the other.

    Of course, one will happen before the other. It may be one, it may be the other one. It has no importance at all.

    Their probability is the probability of two independent events, the product of the two probabilities.

    This is the scenario if they are both neutral, as it seems to be the case in our example.

    Of course, if even on of the two were deleterious, the scenario would be much more catastrophic!

    But there is no need for either of them to be deleterious, or to be cancelled by negative selection, as Moran seems to believe. The probabilities we have computed (the same stated by Behe and accepted by Moran) are the probabilities of the two events when the first one is completely neutral, and is perfectly retained in the individual clone where it happens.

    So, again, Behe is perfectly right: chloroquine resistance cannot be obtained in a “stepwise, sequential route”. Not at all. Because, to be initiated, the pathway requires two independent mutations, not stepwise, not sequential, in any sense.

    So, what is Moran saying? That after the first two independent mutations, which already confer a good level of cholroquine resistance, other optimizing mutations are added by RV+NS?

    Yes, that is true, and so? That changes nothing.

    Behe is right all the same. And Moran is wrong all the same.

    Behe has always spoken of two mutations, two independent events that are necessary to reach chloroquine resistance. His probabilistic considerations are about that scenario. And they are completely right.

    And Moran himself admits that Behe could not know the details in Summer’s paper at the time he wrote his book. Therefore, he could not know that, beyond the two initial independent mutations that he had correctly predicted, a few additional mutations could optimize the function by RV+NS.

    The important conclusion, again, is that the first and basic barrier to the RV+NS algorithm is the complexity of the initial variation which generates the selectable function. That is a completely insurmountable barrier for any function with a minimal complexity.

    Adding a few optimizing mutations is rather easy, for simple functions with a modest continuous functional landscape surrounding them.

    But that optimization quickly reaches a roof.

    Another huge probabilistic barrier lies in the search space ocean which surrounds functional islands, preventing any pathway from one island to a different one, and even between separated islands which implement, with different levels of efficientcy, the same function (the rugged landscape).

  334. 334
    Origenes says:

    On Moran’s claim that Behe asks for two simultaneous mutations. Moran isn’t the one who came up with that ‘criticism’:

    Paul Gross: “Behe assumes simultaneous mutations at two sites in the relevant gene, but there is no such necessity and plenty of evidence that cumulativeness, rather than simultaneity, is the rule. As Nature‘s reviewer (Kenneth R. Miller) notes, ‘It would be difficult to imagine a more breathtaking abuse of statistical genetics.’” (The New Criterion, 2007)

    Jerry Coyne: “What has Behe now found to resurrect his campaign for ID? It’s rather pathetic, really. … Behe requires all of the three or four mutations needed to create such an interaction to arise simultaneously. … If it looks impossible, this is only because of Behe’s bizarre and unrealistic assumption that for a protein-protein interaction to evolve, all mutations must occur simultaneously, because the step-by-step path is not adaptive.” (The New Republic, 2007)

    Nick Matzke: “Here is the flabbergasting line of argument. First, Behe admits that CQR evolves naturally, but contends that it requires a highly improbable simultaneous double mutation, occurring in only 1 in 1020 parasites. … The argument collapses at every step.” (Trends In Ecology and Evolution, 2007)

    Sean Carroll: “Behe makes a new set of explicit claims about the limits of Darwinian evolution, claims that are so poorly conceived and readily dispatched that he has unwittingly done his critics a great favor in stating them. … Behe’s main argument rests on the assertion that two or more simultaneous mutations are required for increases in biochemical complexity and that such changes are, except in rare circumstances, beyond the limit of evolution. .. Examples of cumulative selection changing multiple sites in evolving proteins include … pyrimethamine resistance in malarial parasites — a notable omission given Behe’s extensive discussion of malarial drug resistance. … [T]he argument for design has no scientific leg to stand on.” (Science, 2007)

    Where did they get that from?
    From Behe’s book “the edge of evolution”.
    Which part?
    This part:

    Recall that the odds against getting two necessary, independent mutations are the multiplied odds for getting each mutation individually. What if a problem arose that required a cluster of mutations that was twice as complicated as a CCC? (Let’s call it a double CCC.) For example, what if instead of the several amino acid changes needed for chloroquine resistance in malaria, twice that number were needed? In that case the odds would be that for a CCC times itself. Instead of 10^20 cells to solve the evolutionary problem, we would need 10^40 cells. Workers at the University of Georgia have estimated that about a billion billion trillion (10^30) bacterial cells are formed on the earth each and every year. … If that number has been the same over the entire several-billion-year history of the world, then throughout the course of history there would have been slightly fewer than 10^40 cells, a bit less than we’d expect to need to get a double CCC. The conclusion, then, is that the odds are slightly against even one double CCC showing up by Darwinian processes in the entire course of life on earth.

    [Michael Behe, The Edge of Evolution: The Search for the Limits of Darwinism, p. 135]

    Where does it say “simultaneous”?
    I don’t know.

  335. 335
    gpuccio says:

    Origenes:

    Thank you for the documentation of darwinist errors and misrepresentations. The same bad faith is in all those criticisms.

    Thank you also for quoting Behe’s words exactly (I had not the book readily available to look for it).

    Of course, there is no reference at all to “simultaneous mutations”.

    To quote Paul Gross:

    “It would be difficult to imagine a more breathtaking abuse of statistical genetics”, and, more in general, of statistical concepts, than what we can find in those “criticisms” by Gross himself, Coyne, Matzke, Carroll, and Moran.

    I would expect nothing better from the likes of Coyne and Matzke, if I have to be sincere, but I must confess that I am a little disappointed of Moran. I thought better of him.

    This is really bad reasoning, and probably bad faith reasoning, of the worst kind.

    If these people had just stopped a moment to consider the basics of probability theory, they would have avoided those gross and foolish statements.

    What has simultaneity to do with a product of probabilities? Do you have to toss two coins simultaneously to have 1 probability out of 4 of getting two heads? Isn’t it the same if you toss one coin twice?

    I suppose Coyne and Moran and others are fond of their scenarios where coins are tossed in the air at the same time, and dice are rolled simultaneously: to quote them again, “If it looks impossible, this is only because of their bizarre and unrealistic assumptions”.

    I am really amazed at the arrogance ot those people.

    Behe’s reasoning is right, humble, pertinent, realistic, correct, scientific and in perfect accord with probability theory.

    Behe is a true scientist, and a very good man.

  336. 336
    Origenes says:

    Let’s assume that Larry Moran acted on trusting his colleagues: ‘they all say that Behe assumes simultaneous mutations, so that must be correct.’
    This scenario doesn’t picture Larry as an independent thinker, but it saves him from having argued out of bad faith.

  337. 337
    gpuccio says:

    Origenes:

    OK: let’s assume that. 🙂

  338. 338
    gilthill says:

    Many thanks gpuccio for all your hard work and wonderful writings here ; it really represents an invaluable resource for anyone interested in the evolution debate.
    Regarding the waiting time problem, I am so pleased with your analysis that vindicates so clearly Behe ´s argument!
    On the same topic, did you read the article by Sanford et al entitled « the wainting time problem in a model hominin population »? It is really a must read; indeed, using a different approach than Behe, namely numerical simulation, Sanford et al demonstrate that it is absolutey impossible to go from ape to man with the RV + NS algorithm.

  339. 339
    Origenes says:

    GPuccio @333

    GPuccio: Another huge probabilistic barrier lies in the search space ocean which surrounds functional islands, preventing any pathway from one island to a different one, and even between separated islands which implement, with different levels of efficientcy, the same function (the rugged landscape).

    Axe provides several arguments as to why the search space for proteins is enormous and function is scarce —a summation:

    (1) Proteins are large; for stability, for being able to have important interactions with the substrate some distance away from the place where the actual chemical conversion occurs, for being able to perform simultaneous processes occurring at different sites on the same enzyme
    (see p.3 and p.4) .

    (2) The rarity of functional folds; each type of fold has a unique complex tertiary structure, SCOP classifcation of protein structures currently has 2,008 different structural categories for protein domains, (more see p.5. and p.6).

    (3) No structural modularity; “the highly cooperative nature of protein folding [44] means that stable structure forms all at once in whole chunks—domains—rather than in small pieces. Consequently, self-contained structural modules only become a reality at the domain level, which makes them unhelpful for explaining new folds at that level” — (see p.8 and p.9).

  340. 340
    gpuccio says:

    gilthill:

    Thank you for your kind words and attention! 🙂

    Defending Behe’s argument is really a pleasure and an honour.

    I have not read that paper, but I will do it as soon as possible. I just wonder, how could they have it published? 🙂

    The abstract is already very interesting:

    To establish a string of two nucleotides required on average 84 million years. To establish a string of five nucleotides required on average 2 billion years.

    .

    That is simply what we expect when we transfer Behe’s results about chloroquine resistance to a human scenario!

    And it’s not different from the results of the “waiting for two mutations” paper, by Durret and Schmidt:

    Consistent with recent experimental observations for Drosophila, we find that a few million years is sufficient, but for humans with a much smaller effective population size, this type of change would take >100 million years.

    OK, so Sanders is still optimistic with his 84 million years for two nucleotides. Durret and Schmidt go for >100 million years!

    However, the two estimates are quite similar. It seems that we are in the right order of magnitude, for that specific problem.

    Good luck to all my darwinist friends! 🙂

  341. 341
    gpuccio says:

    Origenes:

    Thank you again for the references and the summary.

    Of course, Axe is perfectly right! 🙂

  342. 342
    gpuccio says:

    To all interested:

    But let’s go back to Natural selection, and to the protein functional space.

    I think we all agree, even oru kind darwinist interlocutors, that RV alone can do little: and yet we have seen that it has the fundamental role, in the algorithm, to find the initial starting function.

    And the initial starting function must always be simple: it’s not important if it is 1, or 2, or 3, or 4 AAs. It’s always simple.

    And the improbability of finding it by RV alone increases esponentially with each new aminocid that is added to the initial functional nucleus.

    Now, just to illustrate the reasoning, let’s refer again to an old friend: ATP synthase.

    In particular, the two alpha and beta chains which, in three copies each, make up the bulk of the F1 subunit.

    You can see a good picture of the whole moolecule here:

    https://en.wikipedia.org/wiki/ATP_synthase#/media/File:Atp_synthase.PNG

    The alpha and beta chians are the part in red and pink, in the lower part of the picture.

    That is exactly the part that binds ATP and pohsphate, then changes its conformation and squeezes that two molecules together, synthesizing ATP and storing in its molecule the biochemical energy that comes from the work of the rest of the molecule. You can see a good video explaining the main concepts here:

    https://www.youtube.com/watch?v=b_cp8MsnZFA

    and another one here:

    https://www.youtube.com/watch?v=39UKSfsc9Z0

    Now, we are discussing here only the sub-part which operates the final function:

    a) binds ADP and phsphate

    b) squeezes them together

    c) releases ATP, retwrning to the original conformation

    Therefore, this part undregoes three conformational changes: those changes are effected by the rotation of a motor, made of the three remaining chains of the F0 subunit, gamma, delta and epsilon, which in turn rotates because it is coupled to the energy derived from the passage of protons from one side of the membrane (the intermembrane space) to the other side (the mithocondrial matrix).

    Now, why are we discussing the alpha-beta part of the F1 subunit?

    Because it is by far the most conserved part of the whole molecule.

    As I have said many times, the alpha and beta chain are two rather long protein chains (553 and 529 AAs), extremely conserved in all forms of life.

    Just to examplify, the alpha chain presents:

    290 identities, 373 positives and 561 bits of homology

    between humans and E. coli.

    The beta chain presents:

    334 identities, 383 positives and 663 bits of homology

    between humans and E. coli.

    So, we can certainly argue that the specific geometry and biochemical functionality of this part of the F1 subunit must be extremely fine.tuned and precise, to make the molecule capable of assuming the three specific conformations that allow it to do what it does.

    More than 1000 aminoacids are necessary to get the structure that is repeated three times in the hexamer. 624 of them have been conserved for billions of years!

    More in next post.

  343. 343
    gpuccio says:

    To all interested:

    Now, let’s try to imagine how sich a perfect machine could originate by RV + NS.

    Just to understand, compare that with Szostak’s ATP binding protein, which, even in its strong version, requires only a few specific aminoacids to work.

    There is the same difference, indeed a much greater difference, than between a formula 1 prototype and a very simple and primitive cart!

    So, how could this big, complex, specific structure originate

    Even if it derived fron some existing structure which did something else (and of which we have absolutely no idea, least of all evidence), there must have been an initial form which could perform the actual funtion, and was then optimized by NS.

    That initial form must have been in the range of pure RV. A few AAs at most.

    So, can you conceive a RV of, say 5 AAs that generates a structure, even if simpler than the one we observe today (and which was however ready almost at the beginning of natural history), which can assume the three sonformations as a result of the rotation of an inner stalk, coupled to the movements of a protoin gradient through a membrane?

    If you can, you have certainly a much greater imagination than I have!

    But that initial 5 AAs random state must be the starting point for a state that, form bacteria to humans, has been based on about 600 conserved aminoacids!

    Now, do you know how specific a target is a 600+ conserved aminoacid sequence?

    We are discussing one state out of 20^600, here.

    I will not insist on that: it should be completely self-evident.

    Now, to reach that state starting from a five AAs initial state generated by RV (which is in itself a huge accomplishment, but you can make it ten aminoacids, I feel generous today!), we should believe that we had about 600 single aminoacid mutations, each of them conferring a definite improvement of the initial ATP synthase activity, IOWs a ladder of 600 events similar to the two or three events we found in the optimization of chloroquine resistance!

    We should believe that, surrounding that initial 10 AAs variation and starting function, there is for some reason not a smal optimization space of a few aminoacids, as we have observed in all real examples, but a huge and continuous optimization space of 600 gradual steps!

    Now, is there any conceptual reason why that idea should be true?

    And is there any empirical evidence that it is true?

    Oh, I realize that I have expressed my challenge again.

    With a specific example, just to help those who want to answer. 🙂

  344. 344
    gpuccio says:

    To all interested:

    Now, what do we know about continuous optimization spaces?

    a) We know that they have been observed in a few simple cases of simple initial function. Almost always of the kind: “beneficial loss of function” in an already existing complex and functional structure.

    b) We know that they are small: a few aminoacids at most, in all observed cases.

    c) We know that the process must be continuous, usually with one single stepwise and beneficail mutations.

    But we also know that the space bewteen initial starting functions is not continuous, not even for the same function. It is a rugged space.

    In the case of chloroquine resistance, we have seen that it is impossible to pass from one of the two possible starting states to the other: a mini rugged landscape from all points of view, in a very simple syste, of two staring states of two AAs each, with one aminoacid in common!

    That is really amazing evidence for a discontinuous space.

    But let’s look to another well documented model: the rugged landscape experiment itself.

    I quote a very important statement in that paper:

    More than one such mountain exists in the fitness landscape of the function for the D2 domain in phage infectivity. The sequence selected finally at the 20th generation has ?=?0.52 but showed no homology to the wild-type D2 domain, which was located around the fitness of the global peak. The two sequences would show significant homology around 52% if they were located on the same mountain. Therefore, they seem to have climbed up different mountains.

    Emphasis mine.

    Now, that’s really amazing. The best sequence they got as a result of RV + NS and the wildtype sequence showed no homology at all!.

    IOWs, their functional islands are completely separated in the ocean of the search space.

    There is no way to pass from one to the other. Because it is too unlikely.

    But also for another good reason.

    The reason is that the “mounts” used even in the paper to represent local functional islands are not mount at all. They are holes.

    The rugged landscape is a landscape made of huge, almost infinite flat surfaces (the vast parts of the search space that are neutral to the function) and a number of distant, completely separated holes.

    Holes are the islands of local optima, where some optimization can be achieved by small stepwise mutations.

    IOWs, the only parts of the system where the functional landscape is in some way continuous.

    If you arrive, by RV, in the small space of one hole, you will most likely fall to the bottom of it (the local optimum). You have almost no chance to climb again out of the hole.

    So, the functional landscape between starting states is completely discontinuous: they are isolated in the search space, and negative selection acts against any chance of going from one to the other, even if we are speaking of starting states for the same function.

  345. 345
    Mung says:

    It certainly appears that the more gpuccio has to say on the matter the less any critic is willing to step up and defend the Darwinian scenario. Maybe gpuccio should say less.

    😀

  346. 346
    gpuccio says:

    Mung:

    “Maybe gpuccio should say less.”

    That’s probably a good idea! 🙂

    However, I would be happy to have someone engage in my challenge.

    I really believe that the non-deconstructability of complex functions remains the strongest, simplest, definitive argumetn against NS and all reasonings based on it, as far as any fucntion with any relevant functional complexity is concerned.

    They can keep all their simple functions and all their precious microevolutionary scenarios, as much as they like: I really invite them to build even the simplest biuological environment with those minimal capabilites.

    If anyone wants to reason in terms of anything that is important in biolpogy, enzymes, biochemical networks, far from equilibrium states, membranes, transports, regulations, and so on, what is immediately and badly needed is:

    hundreds or thousands of different specialized proteins, each of them hundreds of AAs long, each of them based on tons of specific functional information.

    Each of them not desconstructable into simple naturally selectable steps, and of course well beyond the range of any realistic probabilistic scenario involving mere RV.

  347. 347
    Dionisio says:

    Mung @345:

    At least gpuccio asks honest questions. Maybe that’s why he is supported by a distinguished Canadian biochemistry professor who (sometimes) comments here. 🙂

    BTW, did you notice this discussion thread has been visited so many times since October 5?

    Popular Posts (Last 30 Days)

    What are the limits of Natural Selection? An interesting… (2,628)

    Violence is Inherent in Atheist Politics (2,031)

    Of course: Mathematics perpetuates white privilege (1,143)

    Sweeping the Origin of Life Under the Rug (1,012)

    Has the missing matter of our universe finally been found? (850)

  348. 348
    Origenes says:

    GPuccio @320

    All these well known cases can be defined as cases of loss of function mutations which indirectly confer an advantage in a specific selective environment.
    Let’s call this kind of mutations: beneficial loss-of-function mutations.
    The point is very simple: loss of function mutations are very simple, even those which are in a sense “beneficial”. That’s why their “target size” is large. That’s why they are in the range of RV.

    There are many ways to destroy a thing, but only a few ways to build it. You have provided us with many insights. This seems to be an important one among many.

  349. 349
    EugeneS says:

    GP,

    Regarding the estimate (your comment 209), I actually tried to do the derivation myself with pen and paper. What I am getting is roughly 10^43 states available to evolution (and correspondingly, 143 bits of functional info, not 140 bits). The difference stems from one extra order of magnitude that creeps in when assessing the number of days in 5 billion years = 1.825E+13 😉

    Anyway…

    I posted a relevant comment in a different thread. I’d like to copy it here as well.

    The criticism I came across when getting the same ballpark figures is, ok, evolution only had time to forage a tiny fraction of the search space and yet, it could build up all the observed biocomplexity!

    What I want to say, is that the estimate of the (minuscule) fraction of the search space that can be visited by evolutionary random walk is not enough by itself. It must be supported by an estimate of rarity of functional states in the search space (such as the one produced by D. Axe, which is 1 functional polypeptide in every 10^77 on average). These two estimates together are a statistical argument against the “grand show of evolution”, as R. Dawkins put it describing the R. Lenski experiment.

  350. 350
    gpuccio says:

    EugeneS:

    Sorry for answering your interesting comment at #349 so late, I had not seen it!

    “Regarding the estimate (your comment 209), I actually tried to do the derivation myself with pen and paper. What I am getting is roughly 10^43 states available to evolution (and correspondingly, 143 bits of functional info, not 140 bits). The difference stems from one extra order of magnitude that creeps in when assessing the number of days in 5 billion years = 1.825E+13”

    Why? Just to understand. I repeated the computation, and it still gives me 1.825E+12.

    I though the difference was due to the fact that I had used 365 days instead of the more precise 365.25. But even using that, what I get is 1.82625E+12

    “What I want to say, is that the estimate of the (minuscule) fraction of the search space that can be visited by evolutionary random walk is not enough by itself. It must be supported by an estimate of rarity of functional states in the search space (such as the one produced by D. Axe, which is 1 functional polypeptide in every 10^77 on average). These two estimates together are a statistical argument against the “grand show of evolution”, as R. Dawkins put it describing the R. Lenski experiment.”

    You are right. But in all my reasonings I use the functional complexity of proteins as measured by their sequence conservation thorugh long evolutionary times. That is in itself a measure of the target space/search space ratio for that protein, so it is a measure, certainly approximate, of the rarity of its functional state.

    And the results are in good accrod with Axe’s results, which however are a measure of the probability of folding, not of the specific function of each protein.

    As I have often said, the rugged landscape experiment, discussed in detail in this thread, gives very good empirical support to Axe’s results.

    Measuring the functional complexity by sequence conservation through very long evolutionary periods using an universally accepted parameter (the BLAST bitscore), is a very simple, flexible and reliable method to approximately measure the target space/search space ratio for any individual protein. The method, even if differently implemented, is based on the same ideas used by Durston in his fundamental paper:

    Measuring the functional sequence complexity of proteins

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2217542/pdf/1742-4682-4-47.pdf

  351. 351
    Eugene S says:

    GPuccio,

    Thank you very much for your time. I don’t have the derivation in front of me just now. You are probably right, and I am wrong, by the look of it.

    On rarity, what I was saying was that opponents could say: we agree, so small a fraction of the search space that has actually been visited, and nonetheless, we observe all this bio-complexity. We need to have more information to assess evolability (i.e. how much could be reasonably expected of random walk to traverse). Without it, it is more or less guesswork. There are some good ideas though.

    I actually had a chat with someone about the rarity of function in protein sequence space. They pointed me to what they consider as evidence against rarity. I am not qualified to judge that but it would be interesting to hear your opinion.

    The family Buprestidae is among the largest of the beetles, with some 15,000 species known in 450 genera. As far as I understood from our opponent, one of the current explanations is neutral evolution.

    To repeat, this example was put forward as evidence against the rarity of protein functions in sequence space. It appears, there are some very dense clusters of solutions in it which can be traversed by random walk/neutral drift.

    I don’t know what evidence (if at all) they have supporting the claim that “neutral drift did it”. It would be nice to have an expert look into this.

Leave a Reply