Uncommon Descent Serving The Intelligent Design Community

Why would I want my doctor to have studied evolution?

Salvador Cordova
March 12, 2007
Intelligent Design
Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

From Dr. Michael Egnor:

No Nobel prize in medicine has ever been awarded for work in evolutionary biology. In fact, I think it’s safe to say that the only contribution evolution has made to modern medicine is to take it down the horrific road of eugenics, which brought forced sterilization and bodily harm to many thousands of Americans in the early 1900s. That’s a contribution which has brought shame—not advance—to the medical field.

So ‘Why would I want my doctor to have studied evolution?’ I wouldn’t. Evolutionary biology isn’t important to modern medicine. That answer won’t win the ‘Alliance for Science’ prize. It’s just the truth.

MORE

Comments
Comment in filter . . .kairosfocus
March 20, 2007
March
03
Mar
20
20
2007
11:39 PM
11
11
39
PM
PDT
great_ape: "1. Frequencies of the S-allele, which causes sickle-cell anemia in its homozygous state are, as a matter of empirical fact, higher in malaria-prone regions than elsewhere. This is based on real genotype data, explains why African-americans have higher sickle-cell incidents, and I presume no one is questioning this fact here." Yes, I'm questioning it. Where is the data? How well documented is it? great_ape: "3. More than one testable prediction can be made based on Haldane’s hypothesis. I detail them below: 3a. There should be a decreased incidence of infection of malaria for SA carriers *and/or* a decreased incidence of mortality from malaria for SA carriers." Why would sickle-cell anemia prevent malaria? Mosquitos deposit the parasite in the blood system; only a small percentage of the RBC are affected, which the body is constantly attacking and destroying, leaving perfectly healthy RBCs to be infected. Further, if you want to say that the "fact" that the S-allele frequencyis higher in adults than in children, suggesting that this is a "slam dunk" for heterozygote advantage, then understand that at the same time it completely dismantles the case for the S-allele preventing infection. You can't have your cake and eat it too. Oops, another aphorism. Please excuse me. great_ape: "3. They note in their methods, when discussing the fitness of genotypes modelled, that there is “good evidence that malarial mortality is the major contributor to differences in relative fitness between AA and AS where malaria is hyperendemic [Note: referencess Aido et al 2002 and Williams et al 2005]. So please note that the authors of this paper concede as established, one of the major points in contention here." But "good evidence" is not the same as "convincing evidence", or even "proof". Maybe it's just a "hunch" that they have, like Haldane. G-A: "Using a starting allele frequency based on the *current* allele frequencies found in the regions of origin in Africa. They ranged this parameter over a window that included multiple African regions contributing individuals. " So, they were guessing. G-A "This part of the simulation was *consistent* with their model, indicating that the allele frequency fell *as expected* over the 7 generations until 1973-1981, just as Haldane’s hypothesis predicted it should when the positive benefit (malaria resistance) was nullified and only the negative consequences of the S-allele remained." How do you know that we're dealing with "malarial resistance"? Are you guessing? G-A "In the *second* phase of the simulation, they looked at the last generation(s) from 1973-1981 thru 1995-2003. The further decline expected by their model during *this specific time period* was not observed." Isn't this interesting: when they're guessing what the allele frequencies were, and how fast they might decline, the model works. Then when you have two REAL points in time, all of a sudden the model DOESN'T work. Very interesting. G-A "#4. Finally, they correctly suggest that since the S-allele frequency is not naturally falling as expected in recent years, it is a problem that may not take care of itself and more effort will be necessary to diagnose, genetically council couples, etc, to reduce sickle-cell incidence." For me, this is the crux of the matter. If, indeed, the S-allele, because of its heterozygote advantage is being "selected" for, then absent the environmental element which renders the heterozgyote more "fit", the allele, since it is so "unfit" in the homozygote, should disappear. But, we don't see that. There appears to be a "bottom" which, in the absence of data, could have been around for a long time. Now, on the other hand, if the S-allele simply makes malaria less virulent---which, indeed, appears to be the case---then, in the presence of epidemic levels of malaria more of the S-alleles would survive and S-allele frequency would rise, and, in its absence, the frequency would fall. Now, I don't consider this NS, but simple stochastic genetic behavior. What we have is a "defective" hemoglobin molecule that has, as an unintended consequence, some remedial effect in the case of malaria. If, indeed, this were NS at work, then in malaria free areas, over the long run, this allele should vanish---IOW, NS would "punish" it. But you see, this study strongly suggests this isn't the case. The Hardy-Weinberg Law seems to be taking over. G-A "Instead, they dwell on the notion that it is due to some inexplicable new S-allele advantage. Maybe they thought they’d get more funding that way? I couldn’t say." The answer as to why they do this is quite simple, really: they're Darwinists, and that's what Darwinists do, they dream up possibilities--emphasis on the dream part. G-A "In either case, the majority of the timeline is *consistent* with Haldane’s prediction. Thus it can be added two the multiple other predictions that held up. It is only the last generation period that breaks the pattern and, I believe, there is good reason to believe this is because of modern innovation that mitigated the lethality of the S-allele, thereby slowing its decline." What "predictions that held up"? Isn't that what we're discussing? You propose that the data from Africa on infection rates supports Haldane's hypothesis, even when the largest sample shows an almost indistinguishable difference in infection rates between S-allele carriers and non-S carriers, and even when the supposed studies showing that there is a big difference between S-allele levels in children and in adults contradicts this hypothesis (and where are these studies exactly? Have we seen them?) Now you want to say that the study your quoting from supports this theory because of a supposed decline in S-allele carriers ( I say "supposed" since we don't know if the S-allele levels declined in the first generation that African blacks were in malaria free Jamaica, or whether it was spread out over all seven, or, if, in fact, the Africans that came over had low S-allele levels in the first place. This study can't ferret any of that out), but the only part of the study that compares real, known data to other real, known data, doesn't conform to their model---which, I'm sure, is based on "Haldane's hypothesis". So, it is premature (and perhaps presumptous)on your part to conclude that there's 'even more evidence' for Haldane's "guess". What's very clear to me, and should be very clear to all, is that, once again, what Darwinists proclaim as "fact", when looked at in detail, looks suspect, if not wrong. G-A "So when I say that 1973-2005 is a drop in the bucket of time, I mean: a)than it only represents 1-2 *sampling* events in the model. Thus there are only 1-2 events for selection to do its thing. I know that in most allele frequency models, across that timespan, you might even see the exact *opposite* trend compared to what is occurring overall." But you see, there's this problem. If they have "modeled" this population for 9 generations, and the model "fits" the data from the first cohort, then, if you've been correct for 9 generations, why would you be wrong all of a sudden for the very next, "single" generation? Again, we're dealing with real, known data, and that real, known data doesn't want to "fit" the model. You, as well as the authors, don't think the model's wrong. I happen to think it is (as I've stated above). Time will tell. G-A "Frankly, if I were them, I’d just be incredibly grateful that I saw the expected decline from the colonization to the 1973-1983 cohort." How could it possibly be wrong? All the numbers are assumed! If it doesn't work, assume new numbers until it "fits". But, that said, I, too, am grateful for their work because at least now we will have real, longitudinal data to work with instead of assumption heaped upon assumption.PaV
March 20, 2007
March
03
Mar
20
20
2007
10:50 PM
10
10
50
PM
PDT
Hi all, sorry to intrude at the end of such an interesting discussion, but I have been away for a few days, so I had to catch up… I am intrigued by the subject of microevolution, because, although I could well not have problems with it in principle, I would like to understand better what is meant by the term. Indeed, I feel that I could share some of PaV’s problems with microevolutionary scenarios, but I have to make some important distinctions. I have read the previous posts about S haemoglobin and malaria with interest, but I don’t understand well where is the problem. In general, I can agree with most of Great Ape’s summaries in posts 97 and followings, in the sense that: a) I can accept that there is a statistically significant correlation between the S trait and malaria infection, both geographically and as incidence data, although it can be true that further studies can be useful to confirm better the correlation itself. b) It seems, also according to the data cited by kairosfocus, that the interpretation of such a correlation as evidence of an “evolutionary” persistence of the S trait due to selective pressure is less obvious. In other words, the correlation exists, but the causal model proposed by Haldane is not necessarily the best explanation. I want to mention here that a statistical correlation is no evidence, in itself, of a specific causal model. The choice of the model and its empirical evaluation is a methodological problem, and statistics can only provide numerical data in support of the inferences made by the researchers. So, if the negative correlation between the prevalence of the S trait and the incidence of malaria can be evidence of a protective effect of the S trait in malaria infection, it is not necessarily evidence of a selective effect of malaria on the S trait. Indeed, the cited discordances in the Jamaica study are against this second inference (that is, Haldane’s theory). c) Anyway, I can accept Haldane’s theory without problems, if it will be supported by facts. Even if it were so, I still don’t understand well in what sense it should be considered evidence of “microevolution”. I will try to explain better this last point. It seems to me that we can comment on 3 important models which are often cited by Darwinists as examples of “microevolution”, or even of “macroevolution”. Part of this discussion was already done with both great ape and kairosfocus in another thread. The 3 models are: 1) Bacterial resistance to antibiotics, in the (minor) form due to point mutations which modify single structures which are the target of the antibiotic. 2) Mosquito resistance to OP pesticides due to point mutations in Ace1 gene. 3) Human (partial) resistance to malaria due to point mutations in the beta haemoglobin chain. As you can see, the 3 models are very similar, for at least three different reasons: a) A point mutation is the cause of the three conditions. b) The mutation is always a variation in the sense of a “loss of information”, with some more or less severe “fitness cost”. The cost is more difficult to evaluate in bacterial resistance (also because there are very different targets involved), has been shown in mosquito mutations, and is obvious in the human S trait. In other word, in all three models we are dealing with something which can be correctly called a genetic “disease”, and more specifically a mendelian disease (point mutation in a gene). c) The “advantage” provided by the mutation is always related to the loss of information, which “interferes” with a specific, “intelligent” aggression from some agent (antibiotics, pesticides, plasmodia). In other words, the “disease” provides an indirect “escape” from some enemy, because it changes (in a degenerative sense) the target of the aggression. This is clear in all three conditions. Let’s discuss better the S haemoglobin - malaria model, just to slow more in detail my points. S haemoglobin is a variant haemoglobin with a single aminoacid substitution in the beta chain. While many single aminoacid substitutions in the haemoglobin chain do not interfere with the protein function, that is not the case with S haemoglobin. S haemoglobin, indeed, is a very unstable molecule, and it tends to precipitate inside the red blood cell (where the haemoglobin concentration is extremely high), especially in conditions of hypoxia. The instability of the molecule is the cause of the severe disease in the homozygous state, and of the mild disease in the heterozygous state, where the quantity of S haemoglobin is lower, because the other allele is normal. So, the point is: S haemoglobin is a “severely less functional” molecule, it precipitates in the red blood cell and is a cause of the sickle deformation of the cell and of its precocious death, with all the pathological consequences. So, you can see that in no way S haemoglobin can be considered an “evolution” of A haemoglobin. That’s exactly the reason why it confers some protection from malaria infection: although the specific mechanism is not well known, indeed it can easily be argued that the presence of the “diseased” S haemoglobin in the red blood cell is a handicap fro the intracellular cycle of the malaria parasite. In other words, S haemoglobin is so much “less functional” that even parasites do not tolerate it! So, even if it were true that the S trait has been “selected” by the pressure of the malaria infection, where is the “microevolution”? At best, we can speak of “facilitation in persistence of a disease because of a second, competing disease”. What has evolved? What has been selected” Where is the “new information”. Please, note that even if we considered these models as examples of evolution, it would be just “minimal” evolution: always and only point mutations, single bit variations. Obviously, CSI or IC are not at stake here. But even so, I have real difficulties to consider them as examples of “minimal” evolution. To me, they are just disease, like the many other mendelian diseases which we must daily diagnose and, to some extent, cure. Do you really mean that carriers of recessive diseases such as drepanocytosis or thalassemia are the beginning of a new species, or even of a significant positive variation in our species? Why? Please, someone from the Darwinist group (great ape, listen) show me a single case of a really “positive” acquisition, even from a point mutation, and then we can start speaking of minimal or microevolution. And then, show me a model of positive acquisition of function through many (let’s say 100) selectable single bit substitutions, and then we can start speaking of macroevolution. And I mean “start”. Otherwise, we shall always remain at the level of unsupported fantasies…gpuccio
March 20, 2007
March
03
Mar
20
20
2007
05:03 PM
5
05
03
PM
PDT
Natural selection in action in malaria though it is not Darwinian but Intelligent Design micro-evolution http://news.wired.com/dynamic/stories/M/MALARIA_RESISTANT_MOSQUITO?SITE=WIRE&SECTION=HOME&TEMPLATE=DEFAULTjerry
March 20, 2007
March
03
Mar
20
20
2007
11:00 AM
11
11
00
AM
PDT
On point 5 above, that should read given 1-3 and possibly 4.great_ape
March 20, 2007
March
03
Mar
20
20
2007
10:22 AM
10
10
22
AM
PDT
I believe we have indeed returned to a similar impass as before, but I believe an objective observer would have no difficulty judging the matter. Finally, some remarks on your concluding points. 1. 700 horses. Without knowing more about these 699 allelic competitors, such as frequency, fitness, degree of malaria resistance conferred, etc, it is impossible to assess if or how well they would compete with the S-allele. 2. Let us be perfectly clear here; there was no *statistical* debate between myself and PaV. I actually calculated the statistics; PaV used no formal analysis whatsoever other than stating his impression of the numbers. That is not science or anything resembling it. 3] "Supportive of the protectiveness hypothesis, is the observed/inferred declining trends in Jamaica etc, but with asterisks on observed flatness in this case." 3. Please distinguish *where* the curve was flat from where the prediction fit, which is the majority of the time period in question >= 75%. 4. I fail to see the logic of #4. Perhaps you could succintly restate it. 5. I feel #5, given 1-3 and possibly 5, is an assertion that is not consistent with the data presented. 6. Well outside the bounds of this discussion, so I won't comment. In sum, from my perspective, I think only your 700 horses argument has potential, but much information and development would be necessary to give enough substance to it to warrant conclusion #5.great_ape
March 20, 2007
March
03
Mar
20
20
2007
06:56 AM
6
06
56
AM
PDT
Concluding [Provisional] Conclusions: 1] There is reason to believe that there are many varieties of haemoglobin, up to 700. A significant fraction of these are believed to be protective against Malaria. 2] Of these, sickle cell trait and anaemia are believed to be somewhat protective [the statistical debate between Pav and GA shows there is some dispute here and there may be interfering factors]. 3] Supportive of the protectiveness hypothesis, is the observed/inferred declining trends in Jamaica etc, but with asterisks on observed flatness in this case. 4] However, as GA has just pointed out, the trend lines are best seen in light of 1000's of gens. That raises the immediate point that the selective pressure on humanity from malaria may not be as long-term as has been imagined, though there is reason to believe that Malaria has been around as long as we have been around swampy regions. (Indeed, across such a long term span, other variants of haemoglobin would be more likely to win the race . . . YEC advocates may therefore want to follow up this one as a test of just how long humans have been around . . . ) 5] More seriously, the case raises significant questions on the testability and confirmed status of NDT in yet another highlighted example of claimed microevolution. Unfortunately, this is a familiar pattern, from Finch's beaks to peppered moths and so on . . .SCOTCH VERDICT: Case not proven 6] On the underlying problem, NDT-driven Macroevolution, we have not even got close to the 500-bit minimum threshold for generating genetic information from “lucky noise” to begin to test that we have empirical reason to believe that RM + NS can create body plan level biodiversity, including credibly irreducibly complex systems. E.g., not only haemoglobin but the other molecules, then the cell types, tissues and organs and control systems required for animals with a blood based circulatory system. The gaps dominate the evidence in short. Of course, across chance, necessity and agency we know that agency routinely generates such functionally specified, complex and often fine tuned, information ands systems. So, inference to design is a very live issue. GA, that brings us back to the point where we were previously, esp since there has been no onward facilitation by was it Patrick . . . GEM of TKIkairosfocus
March 20, 2007
March
03
Mar
20
20
2007
02:31 AM
2
02
31
AM
PDT
Still having a problem . . . GEMkairosfocus
March 20, 2007
March
03
Mar
20
20
2007
02:30 AM
2
02
30
AM
PDT
Continuing . . . The above got through, partial confirming of h-link hyp . .. 3] Moreover . . . Given the young age at first birth in a lot of Jamaica's population, the 20-30 year separation of the two screenings is also a pretty good proxy for your adult/child decline in frequency of S, in no 97, 3a. So, the flatness trend observed is probably not insignificant, even across the span of only one generation. [I am still pointing to the potential of sampling mothers and their children and using date of mother's birth . . .] So, on looking at your 100, the key point seems to be a bit of a split on the narrower point: S frequency does decline over 7 or so gens, but stabilises over the last 1 – 2. Note too, how they are coming out on the age of Ja's pop base about where I was: 30 * 9 = 270 y. And, I note that mortality from SCA etc is STILL a problem. The 50's – 60's were an early transitional era in Jamaica's health care and infrastructure in general, and on general familiarity with Jamaica's development and health care picture, I am very willing to accept the implication of the authors that the bulk of improvement in both care and resulting mortality statistics due to that etc happened 70's – 80's, based on patterns of infrastructure development in the 50's to early 70's and the sharp rise in penetration of personnel over the later span [especially in the 70's – early 80's], tied to development of education systems and policies to man the infrastructure and carry forward the many health and development projects, some of which my Mom worked on in the Bureau of Health Education late 60's – mid 70's. It is of course the cohort of the 50's – 60's whose births would be showing up in the 70's – 80's, and this second cohort whose children would be showing up in the 90's to 00's. Thus, the adult generation, child generation downward trend effect should be evident in the samples. The authors therefore properly highlight the observed fact that it does not. 4] On generations, samples and noise: You have accepted that the non-overlapping generation approximation is a useful heuristic. That fits with the point that maternal birthdate is relevant as I noted earlier. But, cost is always a factor that leads to simplification in research work. A range of about 15 – 45 years does average to about 30 years, and in my own family's direct line, 6 generations carries me back roughly to turn of C19, so the approximation is valid for my family on average. [But, there are several cases of multigenerational teen mothers in my [fairly large!] wider family. So, at death, one of my grandfathers had Great-Great grandchildren at 95; 5 generations born in 95 years is less than 20 years average.] I am not very impressed by “noise” claims in a context of LARGE samples, taken over half a decade or more. Jamaica has probably 10E4 or better births per year. If there is a trend line, and the model [however proxy-riddled and crude] has been more or less validated [across 7 gens], in a context where there is enough sensitivity in the sampling processes to pick up S- frequency falls between adults and children in a population, then there is enough data to see the onward trend if there is one. The observation is, no, there is no trend statistically significantly different from flat. [And, they are probably well beyond the point where the law of large numbers kicks in.] So, subject to further data, we are looking at . . . [Provisional] Conclusions: 1] There is reason to believe that there are many varieties of haemoglobin, up to 700. A significant fraction of these are believed to be protective against Malaria. 2] Of these, sickle cell trait and anaemia are believed to be somewhat protective [the statistical debate between Pav and GA shows there is some dispute here and there may be interfering factors]. 3] Supportive of the protectiveness hypothesis, is the observed/inferred declining trends in Jamaica etc, but with asterisks on observed flatness in this case. 4] However, as GA has just pointed out, the trend lines are best seen in light of 1000's of gens. That raises the immediate point that the selective pressure on humanity from malaria may not be as long-term as has been imagined, though there is reason to believe that Malaria has been around as long as we have been around swampy regions. (Indeed, across such a long term span, other variants of haemoglobin would be more likely to win the race . . . YEC advocates may therefore want to follow up this one as a test of just how long humans have been around . . . ) 5] More seriously, the case raises significant questions on the testability and confirmed status of NDT in yet another highlighted example of claimed microevolution. Unfortunately, this is a familiar pattern, from Finch's beaks to peppered moths and so on . . .SCOTCH VERDICT: Case not proven 6] On the underlying problem, NDT-driven Macroevolution, we have not even got close to the 500-bit minimum threshold for generating genetic information from “lucky noise” to begin to test that we have empirical reason to believe that RM + NS can create body plan level biodiversity, including credibly irreducibly complex systems. E.g., not only haemoglobin but the other molecules, then the cell types, tissues and organs and control systems required for animals with a blood based circulatory system. The gaps dominate the evidence in short. Of course, across chance, necessity and agency we know that agency routinely generates such functionally specified, complex and often fine tuned, information ands systems. So, inference to design is a very live issue. GA, that brings us back to the point where we were previously, esp since there has been no onward facilitation by was it Patrick . . . GEM of TKIkairosfocus
March 20, 2007
March
03
Mar
20
20
2007
02:28 AM
2
02
28
AM
PDT
GA, PaV and others: Okay. Seems that the first post in the sequence that was blocked yesterday made it overnight . . . [I didn't try to post the other two parts] Also, appreciation GA for pulling the paper [I do not currently have access to EBSCO host etc]. The evident use of infants only, in a context where mothers etc to the 1930's would have naturally been available, is probably a reflection of funding constraints on the health service – a pity, but a reality. [NB: sufficient of cohorts to the 1920's are still available among the aged to get some indications for a multidecadal trend line, but this is of course a clock ticking away at this point. Your point on the relatively low incidence of reported Malaria deaths in Jamaica is also taken on board, though a parish by parish breakdown may have revealed a key pattern – there are several relatively large scale swampy, low-lying areas [Near the Hellshire region, the Black River Morass, Parts of Westmoreland stand out in my mind, and many areas that are not like that, so a more detailed breakdown may be helpful. But, they are using W African rates as proxies . . . easier and cheaper. BTW, there is an implication that low enough selection pressure has no observable effect.] I also note that even with screening of infants, if we are looking at a trend line that is visible on 20 years span, selecting infants based on maternal age at birth in annual cohorts is going to reflect trends too: kids born to 15 yo's vs 45 yo's. Birth certificates in Jamaica (thus a fortiori routinely available data in hospitals and clinics) will have that information. [Note here your citing of HH on how the trends on S are claimed to be visible adult/child in one generation span.] So, there are credible ways to do follow up work on those trend lines. But, money is a constraint again doubtless. Now, on key points to be followed up: 1] Going back on "rare" The Atlanta fact sheet [I am suspecting h-links on the filter problem, so kindly look back at no 65] actually warned SA folks as follows:
Persons who have Sickle Cell Trait should be aware of the following: 1] They are advised against flying in unpressurized aircraft at an altitude above 8,000 feet 2] They are advised against going deep-sea diving or high mountain climbing, due to the possibility of sickling of the red blood cells 3] If they should see blood in his/her urine (hematuria), he/she should see a doctor immediately 4] Persons should let their anesthesiologist and doctors know their condition it they need to have any type of surgery There are over 700 different types of hemoglobin. Many of these serve as a protection against Falciparum Malaria. They are not contagious and can only be passed from parent to child. It is important for you to know your hemoglobin type.
In short, even brief exposure to relatively high altitudes is problematic. That would hold for mot only flying, but going to market in the hills to sell produce, etc. [In British territories, the wealthy tended to live in seaport towns, in Spanish ones, they often preferred to live in higher cooler –- relatively Malaria-free – areas]. Surgery etc are also not very rare, for public health and pop trend concerns. [A seriously deleterious or deadly condition is sufficient of a bad outcome, that even for lowish probabilities, one needs to be warned and cautious.] Third, we are evidently in a multi-horse race, in a context where Malaria is said to be the major selection pressure on the human genome in recent times. That raises the underlying point PaV and I have made, that this is not a two or three horse race, so if pop dynamics have ex hypothesi had time and capacity to play out to select for S, it would also have had time to select for OTHER protective traits that are not so damaging. And, that means the genetic race should have been won by something else, if protective effect is the critical driving force. 2] Thus . . . it seems to me PaV has long since [also relative to the abstract] made the key take-home [or, at least take-to clinic] lesson points above, in 80:
[1] when they compared actual frequencies between two different neo-natal screenings, which took place twenty years apart, that the S-allele did not decrease in frequency, and that the only way that they could get their model’s predicted decrease to match the non-significant decrease of S-allele frequency was by assigning a “recent” (meaning ‘twenty years later’, not in the original testing) higher fitness value to the HbSS homozygotes. . . . . [2] given that there are so many different types of ‘heterozygotes’ for hemoglobin—sickle cell, thalassemia, dust antigen, HLA, etc., etc.—the more sensible interpretation is not that ALL of these confer any kind of ‘heterozygote advantage’, but simply that when it comes to hemoglobin its complex structure permits a number of heterozygotes appearing without lethality. IOW, the recessive homozygotes are not sufficiently lethal as to eliminate them from a population entirely, so there remains a kind of ‘residual’ allele frequency.
Pausing . . .kairosfocus
March 20, 2007
March
03
Mar
20
20
2007
02:23 AM
2
02
23
AM
PDT
"“Rather brief” is therefore credibly a material misrepresentation. So is “ barely time for a generation turnover.“" --kairosfocus We are coming, as you recognized, from two very different perspectives. Models of the sort they are employing typically use *non-overlapping* generations. (I checked, they are indeed simulating 9 non-overlapping generations since the first slaves were transported) That is, each 20-30yrs span represents a single generation turnover. This, of course, is a gross over-simplification of reality. But overlapping generations are a nightmare to model and, from what I understand, non-overlapping generations capture the demographic picture adequately *most* of the time. So when I say that 1973-2005 is a drop in the bucket of time, I mean: a)than it only represents 1-2 *sampling* events in the model. Thus there are only 1-2 events for selection to do its thing. I know that in most allele frequency models, across that timespan, you might even see the exact *opposite* trend compared to what is occurring overall. Their model shows a deterministic trend downwards. But the *empirical* datapoint could be drifting erratically over the same period because if 2 sampling events are what's going on in model-world, and if model-world captures real-world reasonably well, then two sampling events doesn't leave much room to smooth out random noisy bumps in allele frequencies. b)These sorts of allele frequency simulations are typically done over 10s of 1000s of years (if not longer) so that is why I find it a bit odd for them to be thinking on such short timescales. In particular, I find it odd for them to be so comfortable with modeling such short timescales (in human generations; bacteria and drosophila are another story). I think the authors were both creative and bold to try testing the hypothesis this way, I just think they should have been far more skeptical about what they could reasonably determine from the period of 1973-2005. Frankly, if I were them, I'd just be incredibly grateful that I saw the expected decline from the colonization to the 1973-1983 cohort. Yet hats off to them if they have indeed noticed that the decline is flattening out. Though I disagree with their interpretation of *why* it might be flattening out, it is clear to me that more counseling and other intervention is needed if their 1973 thru 2005 results hold up to further scrutiny.great_ape
March 19, 2007
March
03
Mar
19
19
2007
09:43 PM
9
09
43
PM
PDT
(continuing from above) #3 The authors suggest increased transmission of S-allele to offspring may be a contributing factor in the S-allele's persistence and maintenance above and beyond malaria resistance. The effect has been noted in a previous Brazilian study. However, they note that this effect is a minor player. They suggest that *maybe*, with malaria out of the picture, this effect might manifest itself to bolster the S-allele further than it might be expected. This is followed by a disclaimer, indicating that the reviewers thought it was bologna and wished it removed or severely weakened. #4. Finally, they correctly suggest that since the S-allele frequency is not naturally falling as expected in recent years, it is a problem that may not take care of itself and more effort will be necessary to diagnose, genetically council couples, etc, to reduce sickle-cell incidence. In conclusion, I have two issues. 1) I question how much whether noise may have prevented them seeing a decline in S-frequency over the last generation (i.e. between the 1973/1983 and 1995/2003 cohorts). 2) Even if this is correct, I question why they so readily dismiss the possibility that the improved S-allele relative fitness that slowed the S-allele decline in the last generation was not attributable to health care and other modern improvements. Instead, they dwell on the notion that it is due to some inexplicable new S-allele advantage. Maybe they thought they'd get more funding that way? I couldn't say. In either case, the majority of the timeline is *consistent* with Haldane's prediction. Thus it can be added two the multiple other predictions that held up. It is only the last generation period that breaks the pattern and, I believe, there is good reason to believe this is because of modern innovation that mitigated the lethality of the S-allele, thereby slowing its decline. I hope this has brought some clarity.great_ape
March 19, 2007
March
03
Mar
19
19
2007
09:10 PM
9
09
10
PM
PDT
Pardon my numerical numbering issues above... Continuing... The decline in the final generation in the simulation is greater than what is actually observed. They dwell on this point because it is the most incongruent with expectation, and they can get the most "newsworthy" mileage from it. In order to make the data *fit*, they have to increase the fitness of the S-allele during the *last* generation. They offer several possible explanations, and ultimately give many caveats, which is as it should be. #1. The model is wrong. (too simple, parameters off, etc.) They clearly do are not favorably inclined towards this explanation. #2. (my personal favorite) Better healthcare has rendered the S-allele less lethal in the modern period (post-1960 or so), reducing the rate of decline of S alleles in the last generation. On one interpretation, this is why their "retrodictive" increasing of the S-fitness parameter for the last generation(s) makes it more consistent with he data. --The authors don't buy this explanation, however, because they don't think sickle-cell lethality was *that* much improved in the modern era. They offer weak argument IMO, however, for why this should be believed. Specifically, they cite the incredible improvements in survival in the 70s thru the 90s as evidence things could *not* have been improving much in the 50s and 60s. This, to me, seems a poor reason to dismiss possibility #2 when no sickle-cell mortality data from the 1950s and 1960s were examined examined. [pause.]great_ape
March 19, 2007
March
03
Mar
19
19
2007
08:50 PM
8
08
50
PM
PDT
3c. Okay, I have the details of the Hanchard et al. 2006 paper mentioned above. A few pertinent points: 1. The allele frequency data for the study is from *neonatal* screening. As such, the 1973 cohort begins with individuals born in 1973 and ends with those born in 1981. The same applies to the 1995-2005 cohort. 2. As I anticipated from the abstract, there are *two* components to the simulation. One is the prediction of allele frequency decline from 3. They note in their methods, when discussing the fitness of genotypes modelled, that there is "good evidence that malarial mortality is the major contributor to differences in relative fitness between AA and AS where malaria is hyperendemic [Note: referencess Aido et al 2002 and Williams et al 2005]. So please note that the authors of this paper concede as established, one of the major points in contention here. 3. When DDT was introduced to Jamaica is, by and large, irrelevant here. The authors show empirical data indicating that the mean death-rate for mortality via malaria from 1886-1927 was 40.8 per 100,000 in Jamaica. This is compared to mean rates of 500-1000 per 100,000 *currently* in the regions of Africa from which the populations came. In effect, the selection from malaria in Jamaica was negligible compared to Africa. This is a premise of the study. After the population relocated, the selective effect deriving from malaria was effectively removed. 4. As I anticipated from the abstract, there are *two* components of the simulation. In the first component, allele frequencies of the *first* cohort 1973-1981 are compared to what they would be after 7 roughly 30-year generations since relocation of the population to Jamaica. Using a starting allele frequency based on the *current* allele frequencies found in the regions of origin in Africa. They ranged this parameter over a window that included multiple African regions contributing individuals. They then compared how the allele frequency was predicted to decline until the 1973-1981 period with the *actual* 1973-1981 genotype data. This part of the simulation was *consistent* with their model, indicating that the allele frequency fell *as expected* over the 7 generations until 1973-1981, just as Haldane's hypothesis predicted it should when the positive benefit (malaria resistance) was nullified and only the negative consequences of the S-allele remained. In the *second* phase of the simulation, they looked at the last generation(s) from 1973-1981 thru 1995-2003. The further decline expected by their model during *this specific time period* was not observed. The dwell on this point considerably, as I detail below. [pause here...]great_ape
March 19, 2007
March
03
Mar
19
19
2007
08:31 PM
8
08
31
PM
PDT
CORRECTION: "...provided by the S-allele in its homozygous state." --me That should be heterozygous state...great_ape
March 19, 2007
March
03
Mar
19
19
2007
07:55 PM
7
07
55
PM
PDT
kairosfocus, The relevant historical details concerning jamaica are much appreciated. I would like, for a moment, to take a step back and assess what is and is not in agreement here. I am beginning to feel as though we are "losing the forest for the trees" in this discussion and some perspective is necessary. If there is a point listed below which is not accepted, please indicate which. 1. Frequencies of the S-allele, which causes sickle-cell anemia in its homozygous state are, as a matter of empirical fact, higher in malaria-prone regions than elsewhere. This is based on real genotype data, explains why African-americans have higher sickle-cell incidents, and I presume no one is questioning this fact here. 2. Haldane proposed, well before it could be empirically tested, that the reason for the high frequency of the S-allele in these populations was due to a protective benefit against malaria provided by the S-allele in its homozygous state. 3. More than one testable prediction can be made based on Haldane's hypothesis. I detail them below: 3a. There should be a decreased incidence of infection of malaria for SA carriers *and/or* a decreased incidence of mortality from malaria for SA carriers. 3b. If a population in a malaria-endemic area were to relocate to a non-malaria region, the S-allele frequencies would decline. Restated, if the selective pressure preserving the S-allele were removed, the deleterious nature of the SS homozygote would cause the allele frequency of S to decline in the population. Tests of empirical evidence thus far: --> 3a. Consistent with this hypothesis, statistically, is the data PaV first linked to. All populations show a lower incidence of infection for S-containing genotypes than for non-S phenotypes. This represents a comfortably significant statistical effect, indicating *something* is going on between the variables of infection and genotype. -->3a (also). Other studies, referenced by HodorH, indicate that the genotype frequencies for those containing S are *higher* in adults than in children. This is straightforward and powerful evidence for the removal of non-S-allele containing individuals from the population in regions with endemic malaria. --3b. Concerning evidence for the allele decline after a population relocation, I will continue in another post below, as it is more complex. In sum, so far, several empirical tests have failed to exclude Haldane's evolutionary (or microevolutionary) hypothesis based on predictions directly resulting from this hypothesis. This is a pretty good case so far. This is how science works.great_ape
March 19, 2007
March
03
Mar
19
19
2007
07:54 PM
7
07
54
PM
PDT
"G-A, he made an argument that took three posts and touched on numerous points. It’s neither fair nor accurate to say he’s arguing by aphorism." --tribune I was responding to not only his use of aphorism but also to PaV's above where he invoked Twain's and others multiple times. I have, finally, gained access to the full paper that most of this discussion has been based on, and I'll be sending another post discussing it this evening.great_ape
March 19, 2007
March
03
Mar
19
19
2007
04:13 PM
4
04
13
PM
PDT
G-A --There are half-truths, blatant lies and statistricks. That’s because “while figgers don’t lie, liars can figger.” –kairosfocus . . . If I could make scientific arguments by simply repeatedly invoking aphorisms, it would make my professional life much simpler. G-A, he made an argument that took three posts and touched on numerous points. It's neither fair nor accurate to say he's arguing by aphorism.tribune7
March 19, 2007
March
03
Mar
19
19
2007
01:41 PM
1
01
41
PM
PDT
GA: You may be right on how "generation" is defined in your field, but I am speaking from direct observation of my homeland, and as one whose work often requires analysis of empirical data. So, I must insist: given that the reproductive age is ~ 12 - 45, sadly on the lower end [part of Jamaica's major social problems], the samples will have naturally captured mothers born from the 1930's on up to in the case of 2003, the edge of the 1990s, and will have also captured 3+ generations of births, up to 2003. If your technical defns cannot capture that factor they are based on wrong assumptions, period. If I were looking at trend lines on genetic patterns, I would be very interested in birth dates and generations, not just calendar time when the samples were taken, and in mothers not just their kids. So, I repeat, the samples from 1973 to 2003 inclusive credibly capture a range of births from the 1930s to the 2000's, if infants and their mothers are in the screen [which is what would be clinically feasible in Jamaica]. Thus, 3+ biological generations will be reflected in the data. Birth certificates and associated public data [there is but one central islandwide records office that compiles such from parishes, districts and hospitals] would allow a cross-check too. (With enough time and money, patterns in families could be traced . . . thanks to a gift of the Mormon church on genealogy stuff.) I would then split the samples by cohorts on birth timelines back to the 1930's and correlate the genetic data with that, giving trendlines from the 1930s to the 2000s, maybe clumped decadally if the overall sample sizes are too small for annual or half decade clustering. Maybe you have a better answer but that is what looks reasonable to me. (Perhaps you can come up with a better technique that gives better data from these "large" samples, but I would accept nothing less than that much information squeezed out of the easily available and cross-checkable data.) Thus, I can see that the two snapshots credibly can capture trends from the time when Malaria would have likely been still a problem, up to the past half century in which it has for sure been off the PH picture. H'mm: I would want med students to know how to creatively squeeze information out of data too. Trust this helps GEM of TKIkairosfocus
March 19, 2007
March
03
Mar
19
19
2007
10:48 AM
10
10
48
AM
PDT
PaV and GA: 1] DDT made a big difference everywhere. So did major public health initiatives. Until that happened the Caribbean was a tropical deathtrap, not a tourist paradise. (I recall the plaques in St Michael's Cathedral B'dos to the UK regiments stationed there -- they tended to lose a LOT even when no war was on. Disease.] 2] Given the range of reproductive age, ~ 12 - 45, the 20 year apart samples capture 3 + generations of births. [Here I am looking at both mothers and their infants.] That takes us back to the 1930s, which I think we can reasonably accept was malarial, pending more specific data. [Guyana still is in remote parts, to my knowledge.] 3] Slave average life expectancy on arrival was ~ 5 y in C18, and the trade was stopped after about 1807 - 8. So the genetic base of the population is about 250 years old on the outside. 70 years is almost 1/3 of that. (Slaves did not come from any one area.) 4] Sickle Cell is a significant and fairly common public health problem in Ja, and has long been so. So the incidence is on the evidence, high. 5] The key point from the first paper was that it has been flat between the 70's-80s and the 90s-00s, not falling as expected from the same model that matched the level in the first case. This one suggests a mechanism for the unexpected result. That mechanism and the associated context are where the challenges lie, and they point out that med students need a lot more than just the simple story. GEM of TKIkairosfocus
March 19, 2007
March
03
Mar
19
19
2007
10:07 AM
10
10
07
AM
PDT
#1 "1] Reproductive age range is about 12 - 45 in Ja so three plus generations and a span from 1930s on is in the screened population, not just 20 years, and one gen." --kf Matters not. In population genetics terms this is roughly a generation and a drop in the bucket of time. This is a technical issue so unless you have knowledge/training in this that can explain why what I was taught is wrong, I suggest you trust me here. #2. "and that in a post malaria pop that should have a down trend...." The first cohort did show a downward trend, presumably since the colonization. I admittedly don't know much about Jamaican history, but I will try to get the pdf to verify my interpretation. #3 & 4. "The Atlanta brochure, from a reputable source, highlights that There are over 700 different types of hemoglobin. Many of these serve as a protection against Falciparum Malaria." This is a good point, and *could* be relevant, depending on the frequencies of the alleles in question *and* whether or not they are alleles at the same gene locus. There are numerous variations across a population but many are very low frequencies. In terms of your horse race analogy, they are hamsters. 5] The same brochure warns SCT victims against atm press equiv to 8 k ft or more, and surgery. Those are not “rare” issues. --kf In most malaria-infected regions, I don't imagine exposure to heights over 8000 feet is a frequent occurence. I could be wrong.great_ape
March 19, 2007
March
03
Mar
19
19
2007
09:45 AM
9
09
45
AM
PDT
"I didn’t get the impression that a study was done from ‘colonization’ until the 1973-76 time period. How would they have ascertained the allele frequency of those original populations?" --PaV I assume that this longer term study was done because they say that the early cohort's data *is* consistent with the model. But what was the model of? I suspect they used the current allele statistics in the regions affected by malaria from which jamaican's imigrated as the starting frequencies. Then studied the time period following. I will try to hunt down a pdf of the paper to confirm this.great_ape
March 19, 2007
March
03
Mar
19
19
2007
09:26 AM
9
09
26
AM
PDT
great_ape: "Yes, their predictions failed for the specific and rather brief period from the 70s to the 90s. They did not fail for the larger span of history since the initial colonization." I didn't get the impression that a study was done from 'colonization' until the 1973-76 time period. How would they have ascertained the allele frequency of those original populations? And if their model showed that over a twenty year time period the S-allele didn't decline quickly enough, then going back in time, starting with the 1973-76 levels, it seems logical that their model would have had to start with unusually high levels of the S-allele in order to arrive at the 1973 levels. As far as malaria, has the Caribbean been free of malaria for all these centuries, or for only the last 50 years or so? I'm not sure, but it might have been only with the advent of DDT that they were able to get rid of the mosquitoes. As to this particular paper, I think it's a follow-up paper. My impression is that they took the numbers from the first time perios, applied a model, and calculated what the second time period numbers should be and found they were higher than the model predicted. Now, as a follow-up, in this paper they're saying the discrepancy can be accounted for by assuming a higher SS fitness, which, as I said before, is 'retrodicting'.PaV
March 19, 2007
March
03
Mar
19
19
2007
07:14 AM
7
07
14
AM
PDT
Got that through.kairosfocus
March 19, 2007
March
03
Mar
19
19
2007
05:30 AM
5
05
30
AM
PDT
Okay: I think I may get through some real brief points without details etc . . . 1] Reproductive age range is about 12 - 45 in Ja so three plus generations and a span from 1930s on is in the screened population, not just 20 years, and one gen. 2] The peer reviewed material was good enough to pick up a trend as predicted but the trend was not observed, and that in a post malaria pop that should have a down trend. Data was not sig different from flat, leading to suggested clinical interventions. 3] The Atlanta brochure, from a reputable source, highlights that There are over 700 different types of hemoglobin. Many of these serve as a protection against Falciparum Malaria. 4] That means that the number of horses in the race is a relevant issue, not just dragged in. Second to last beats last but does not necessarily come first! 5] The same brochure warns SCT victims against atm press equiv to 8 k ft or more, and surgery. Those are not "rare" issues. So there is more than the just so story that med students should know about. GEM of TKIkairosfocus
March 19, 2007
March
03
Mar
19
19
2007
05:30 AM
5
05
30
AM
PDT
Filter problems . . .kairosfocus
March 19, 2007
March
03
Mar
19
19
2007
05:11 AM
5
05
11
AM
PDT
Following up: 1] GA: their predictions failed for the specific and rather brief period from the 70s to the 90s . . . . a longterm trend (colonization to now) is much easier to observe than one over a couple decades . . . . ~2 decades. That’s barely time for a generation turnover. First, the researchers used the evolutionary population scenario to project trends across the period in question,and found statistically significant results that contradicted the predictions on the trend line: it was flat not trending down. Further to this, the studies were conducted on screenings in the 1970s – 80s and 1990s – 2000s. The findings withstood peer review, well enough to be published. Moreover, mothers giving birth to children circa 1973 would have been born ~ 15 - 30+ or so years earlier, probably during the end of the era in which Malaria was a problem. (Here based on the spreading of DDT, but reckoning that mosquito control before that would have had impacts.) This also immediately implies that the data span a 50+-year period, which is an appreciable fraction of the history of the dominant African population of Jamaica. For, in C18, the typical expected incremental slave lifespan in Jamaica was reportedly ~ 5 years, and the trade was officially ended in 1807 – 8. So, we are looking at maybe 250 years, for the importation of the genetic bulk of the population. Thus, we are looking at a snapshot of the population dynamics over about 20 - 25% of the existence of that population in Jamaica, and starting at the time when we have the ending of the malarial era ranging into something like three generations of births in the post Malaria era, as [given the unfortunate patterns of reproduction that massively prevail] many of the women involved would be unwed teen mothers. (Indeed, I should be saying 12 – 14 years or so at the lower end, but I believe that this is significantly less than 5%. These are LARGE SCALE, routine screenings, where there are probably ~ 10's of thousands of births per year. 5% of 10^4 = 500. If I am an order of magnitude high, that would still capture enough to be within credible statistics. E.g. 8 * 50 = 400, so the short cycle tail of the distribution should include enough to validate the claim that three generations of births are significantly represented, maybe more, as women in their 40s are often enough still in the reproductive population in Jamaica. So, 1973 – 40 = 1933 is also a relevant number to the span of the study.) “Rather brief” is therefore credibly a material misrepresentation. So is “ barely time for a generation turnover.“ 2] They did not fail for the larger span of history since the initial colonization . . . The numbers at the beginning of the period were generally consistent with expected incidence of ACT etc, but the material point is that they did not trend down as ALSO predicted y the models. Now, as this is a recent study [2006 pub date], one wishes for corroboration through other studies in Jamaica and elsewhere, but this is the first step to empirical disconfirmation. It is sufficiently strong a result that the authors make a clinical recommendation designed to foster voluntary action by the counselled to shift the trend. 3] Lethality . . . is only part of the picture; there is also loss in overall reproductive fitness. Precisely, and that is what I pointed out in remarks above. Namely there is reason to believe that heterozygous and homozygous presence of S is harmful, but not sufficiently so to trigger an elimination from the population. Then when we incorporate the rest of the 700 horses, we see that the pictureisa log more complex than the just so story makes out. And that is what med students and the public need to know. 4] we are not only trying to explain why the S-allele *persists*, but also why it persists at such *high frequencies* in those regions plagued with malaria as opposed to other regions. Here is what the peer reviewed study – I hope THIS link is stable -- actually found:
The high frequency of the sickle allele in some parts of Africa is understood to be a consequence of high malarial endemicity. One corollary of this is that the sickle allele frequency should be declining in populations of African ancestry that are no longer exposed to malaria. We have previously shown that there has been no change in sickle allele frequency in malaria-free Jamaica between two large-scale neonatal screening exercises conducted in 1973-1981 and 1995-2003 . . . We found that although model predictions were broadly consistent with observed values in the 1973-1981 cohort, the predicted change in allele frequency between the two cohorts was larger than the observed, nonsignificant, reduction. Close agreement between predicted and observed values was only achieved by simulating a recent, marked increase in HbSS fitness. Thus, the "unexpected" persistence of the sickle allele in Jamaica may reflect the fact that the actual fitness among SS individuals is higher than that previously realized.
Thus the study was, on a peer reviewed assessment basis, good enough to discriminate significant and non-significant trends over the span. The result was, within statistical significance, flat – and flat in a POST-MALARIA, overwhelmingly African derived population. Pausing . . . GEM of TKIkairosfocus
March 19, 2007
March
03
Mar
19
19
2007
05:08 AM
5
05
08
AM
PDT
"There was plainly a base in the population large enough to give meaningful numbers...." Large numbers, yes, but meaningful? Recall that the statistical question involves a trend over time. And nothing is plain or obvious here about the analysis. Bias in the "sampling" of next generation's alleles due to natural selection will be both subtle and subject to a number of noise sources, and it is not straightforward to predict how fast this should happen, or, more importantly, what criteria are necessary to *detect* it happening. I simply question the wisdom behind the authors' interpretation of their results. Did they, for instance, examine random intervals during the longer model period and ask how often they find periods of ~20-30 years where *no* significant decrease was predicted even though the overall (longterm) decrease is clear? (I'm assuming the model was a stochastic computer simulation). I bet they would find numerous time intervals where, by chance, no decrease in allele frequency was observed in the simulation). Yet they choose to invoke another parameter, a novel and unknown genetic benefit, along with a *change* in what's going on biologically in recent decades, to explain the absence of significance over this short time period. Very questionable reasoning IMO. “Haemoglobin S could have a protective role against malaria but evidence is sparse and the operating mechanisms are poorly known.” Once more, quoting an expression of doubt from a single paper from several years ago, particularly when the paper goes on to *show* a quantifiable protective effect of the S-allele across age groups is highly questionable. "Observe the overall context: there is a multi horse race, it is not just on SS, SA, AA. Thus,t he force of PaV’s remark in 80 and my earlier remarks in 64 – 65." --kairosfocus We are specifically discussing the fate of sickle-cell alleles and their relationship to malaria. Certainly there is additional biological complexity, but you merely drag it in without giving any clear indication of how they directly impact the issue at hand. Let's revisit something in your earlier post, where you suggest SA genotype is too deleterious and should be eliminated in such a horse race. You neglected to highlight this particular passage, so I will: "Nevertheless, under unusual circumstances serious morbidity or mortality can result from complications related to polymerization of deoxy-hemoglobin S." Note the part about unusual circumstances. I read unusual circumstances as being particular rare. Once again, it is important to read the entirety of the material you're referring and comprehend it. Do not simply highlight the parts that seem agreeable. The doctors in the study you referenced were justifiably discussing a real medical *possibility* concerning SA that the medical community should be aware of. They specifically note, however, that the circumstances under which SA complications arise are *unusual*. This indicates that your argument that the SA genotype should have lost the race or been trampled by horses or whatever is without substance.great_ape
March 18, 2007
March
03
Mar
18
18
2007
10:03 PM
10
10
03
PM
PDT
"They thus render it a case of “retrodicting”, and not “predicting”—since their “predictions” failed!" -PaV Yes, their predictions failed for the specific and rather brief period from the 70s to the 90s. They did not fail for the larger span of history since the initial colonization. "IOW, the recessive homozygotes are not sufficiently lethal as to eliminate them from a population entirely, so there remains a kind of ‘residual’ allele frequency." --PaV We are specifically discussing sickle-cell alleles and genotypes. Lethality, remember, is only part of the picture; there is also loss in overall reproductive fitness. The fact that homozygous recessive sickle-cell genotypes are less reproductively successful cannot be ignored. The frequency of the S-allele, barring other considerations, ultimately, should go down. How fast depends on numerous factors. The point you're both failing to appreciate is that a longterm trend (colonization to now) is much easier to observe than one over a couple decades. It was the one over the couple of decades that failed to fit their model. Only then do they "retrodict" and insert a new parameter to explain *why* it failed. In my opinion, their retrodiction was a gross error because you can't expect to see a significant decrease in allele frequency over those ~2 decades. That's barely time for a generation turnover. Once you factor in noise, etc, I have no idea why they would even *think* they'd see a decrease here. And then, the icing on the cake is that, to make their new theory work, they have to assume the magical S-benefit is only *recent*. That makes the second part of the paper and all the talk about "no natural reduction" even sillier. Also, we are not only trying to explain why the S-allele *persists*, but also why it persists at such *high frequencies* in those regions plagued with malaria as opposed to other regions. This is an important point that your "just lingering and not very harmful" theory does not adequately address.great_ape
March 18, 2007
March
03
Mar
18
18
2007
09:10 PM
9
09
10
PM
PDT
PS: I add that from the "predicted but not observed declines in Jamaica" study, there is also an issue of HOMOZYGOTE SS viability thus "protective" effect, or at least sufficient non-lethality to survive and persist in the population. [Note too, the reference to counselling, i.e. to hoped-for artificial selection; aka microevolution by intelligent design. H'mm, isn't that what Darwin used as a "substitute" for NS in his own studies and reported on in Origin?] This point on SS, too, does not fit the simple just so story. Let us note: that empirical datum was sufficiently newsworthy among students of haematology to be publishable in a major peer-reviewed journal in 2006. GEM of TKIkairosfocus
March 18, 2007
March
03
Mar
18
18
2007
02:39 AM
2
02
39
AM
PDT
1 2 3 4 5

Leave a Reply