Dan Graur is on a crusade against ENCODE – the NIH project that is showing that, at least in general, the entire genome is functional. Unfortunately, his commitment to the neo-Darwinian view of mutations is blinding him.
Graur recently published a paper in the journal Genome Biology and Evolution saying that the upper limit on the functional fraction of the human genome is 25%.
This paper represents one of the worst instances of how the Modern Synthesis (colloquially known as neo-Darwinism) can blind someone to thinking, and why the concept of Evolutionary Teleonomy is desperately needed in evolutionary biology.
Graur’s Campaign against ENCODE
Graur has been a long-time critic of encode. In fact, Graur has been on record as saying that, essentially, if ENCODE is right, evolution is wrong (his earlier paper criticizing ENCODE for being an “evolution-free gospel” can be found here). At his personal website (very bottom), Graur shows a picture of his baby granddaughter, where Graur’s own caption for the photo is “My granddaughter, Lilla Keshet Graur, gives ENCODE the finger”.
Salvador Cordova has been an active critic of Graur’s attacks on ENCODE. You can find Sal’s previous criticisms of Graur in these links: Gambler’s Epistemology, ALUs, Introns, LncRNAs, and Pseudogenes, and Non-DNA Inheritance.
In the latest round, Graur has been doubling down on his insistence that ENCODE and evolution are at odds, and that the equations of population genetics prove that the genome must be almost entirely non-functional. We reported on Graur’s findings last week based on his slideshow presentation of the idea. This week, since his paper was published, I thought I would go a little deeper into exactly where Graur goes wrong.
Before I get into the meat of the criticism, I should point out that if you agree 100% with Graur’s applications of population genetics, you could not realistically believe in evolution. You would have to agree with John Sanford’s Genetic Entropy concept – that our genomes are not evolving, but are rather in a decaying stasis. Since Graur is obviously never going to agree with that, he is forced to instead simply ignore the problem of the arrival of the fittest, and just assume that it happens. But, once it arrives, it devolves to mostly junk. That’s an interesting thought on its own, but it is not the focus of this essay.
The problem we are going to focus on, is that Graur’s analysis, though seemingly based on empirical findings, is held together by the worst of modern evolutionary thinking – an assumption masquerading as data.
The Modern Synthesis: The Paradigm That Jumped the Shark
When I was newer to the debate, I had always assumed that the concept of “random mutations” was born out by experiment. That we knew for a fact, fact, fact that mutations are mere haphazard events in the cell. I had assumed that since this was emphasized, it must have been found empirically to be true, and we were just debating whether or not it could lead to evolution in the long term.
I was then quite floored when, over the years, I discovered that the idea of random mutations was a holdover from a pre-DNA view of inheritance. It was kept merely because when we discovered DNA we hadn’t yet developed another idea, and it fit in well with the modern synthesis view of mutations (i.e., neo-Darwinism).
Graur’s paper keeps this up par excellence. In his description of the model he is using, he is assuming “that the probability of a mutation occurring in a certain region of the genome is independent of the functionality or lack of functionality of the region in which the mutation arises.” He cites Lederberg and Luria-Delbruck for this. Both of these are pre-DNA papers, and I have shown previously the problem of using these papers to justify that idea of mutation.
This is not a side-issue, in fact, it is the cornerstone of his calculations. Without this assumption, you actually cannot calculate the things he is trying to calculate. The only way that the statistics he is using are combinable is if you assume that there is no prior coordination between mutability and fitness.
As an example, when he is estimating the deleterious mutation rate of missense mutations, he uses a calculation of the total fraction of possible missense mutations and their effect on function (the paper Graur cited is here). This paper is not looking at the actual mutations that organisms undergo, but rather look at all the ways in which mutations can be induced, and seeing how many of them are deleterious.
This is fallacious, unless you accept the neo-Darwinian position that mutations are not geared towards fitness. If you don’t, then these calculations are simply non-sensical.
The Evidence for Fitness-Directed Mutations
As has been shown by others, it appears that the genome is marked in such a way as to point mutation mechanisms to the locations where mutations are helpful (see this, for instance, and this paper showing evidence of coordination of mutations between sites).
This information based on is then combined with empirical data on epistasis. The measured epistasis rate is based on actual epistasis identified among populations that have bred together. Again, this data can only be reasonably combined if the neo-Darwinian picture of mutations is true. You can’t if it is false.
There are other problems with the paper. For instance, it presumes that any mutation that doesn’t affect selectability is essentially non-functional (that is one of the reasons for my recent post on this subject).
Going Forward: Evolutionary Teleonomy as a Replacement Paradigm
So what should we do? No field of study can operate without paradigms. The question is whether the paradigms you are using are moving you forward or inhibiting your progress. Assumptions about the nature of mutations have been hindering progress in evolutionary theory for decades.
My suggestion? Move to evolutionary teleonomy as the background paradigm for evolutionary biology. It comports much better with the data, and doesn’t leave people making ridiculous assumptions about the relationship between mutation and fitness. Biological processes are inherently teleonomical on every scale. Assuming that at the level of mutation biology suddenly becomes non-teleonomic is one of the big mistakes (or even catastrophes) of modern intellectual history.