Friday, August 22, 2008

Response to a comment on my fossil paper and to the recent Cell paper on Neanderthals

leigh van valen commented at Nature precedings on my fossil paper

"There is ample evidence for a variable clock. Unless the divergence from constancy here is greater than what is found elsewhere, the paper unfortunately adds nothing useful."

My response below:

Thank for you this opportunity to address a common mistaken reaction to my paper. I will also here respond to the recent Cell paper on Neanderthal complete mtDNA that has provided independent confirmation of my paper and show that your interpretation of my result is invalid.

Indeed, nearly all the evidence is for a variable clock. I believe that the original molecular clock hypothesis claiming similar mutation rates among different species is largely false. I am like most classical evolution biologists of the 1960s who consider the idea of a constant clock ‘unthinkable’. The tide has now turned that even molecular evolutionary biologists have given up on the idea of a constant clock, faced with mounting evidence of variable clock accumulated after the 1980s.

But most people, when criticizing the constant clock idea, ignored or were simply ignorant of the original value of this idea as a speculative interpretation of the most remarkable result of molecular evolution, the genetic equidistance result first reported by Margoliash in 1963. This result can be restated as the clock: different species have similar mutation rate. This result is extremely robust and universal but has been ignored for 45 years. What is being publicized is not the result itself but the clock interpretation. Even most people within the molecular clock field are not aware how universal this result is and most biologists are completely ignorant of it. Most biologists (greater than 99% based on my informal polling) would guess incorrectly that frogs, for any given protein, are closer to fishes than humans are. The correct answer is that frogs and humans are approximately equidistant to fishes in sequence identity. Humans in fact are slightly closer to fishes than frogs are. I did not know about this genetic equidistance result until a few years ago when I independently rediscovered this result. I was really surprised to find this and soon realized that the clock interpretation of this result is merely a tautology rather than a real explanation. I have now come up with a real scientific explanation with the ‘maximum genetic diversity’ (MGD) hypothesis as you can find in my papers posted here.

So, you see that there are evidence (the genetic equidistance result) that are seemingly consistent with a constant clock, as well as ample evidence against a constant clock. How can a hypothesis be both correct (consistent with the equidistance result) and wrong (inconsistent with evidence for a variable clock)? Either there is a clock or there is not. Only one can be true, constant or variable clock. But what happens in the field is that people use the clock idea to explain one part of reality (the genetic equidistance result) and use the negation of this idea to explain another part of reality (ample evidence of variable clock). This is self-serving and has rendered the clock idea non-testable or non-falsifiable. In my view, the clock interpretation of the genetic equidistance result is dead wrong and the result has a very different explanation. The clock idea is true in only very limited scope, for truly neutral sequences during divergence of two very similar or identical organisms. But most amino acid changes between two distinct species are in fact not neutral for one of the two diverging species and cannot be explained by the clock idea.

Let me explain this. If we can create a yeast cell and human being by using identical genes for their shared homologs and let the two organisms diverge for an infinite amount of time or about 1 billion years, a gene in yeast would have changed a lot to a maximum of, say 50%, while its homolog in human would have changed very little, say less than 1%. Any more changes than 50% would be lethal to yeast and any more changes than 1% would be lethal to humans. The reason that yeast can change much more than human is because a gene in human encounters far more functional constraints than its homolog in yeast. Thus the genetic distance between human and yeast is mainly determined by the mutations in yeast. In this case, the 50% change in yeast would account for the genetic distance of 50% identity between human and yeast, as well as 50% identity in within species distance in yeast. The different residues between yeast and human would be neutral changes only for the yeast but not for human. The mistake of the modern evolution theory is to assume that these changes are neutral for both yeast and human. But the theory is correct only for two diverging species that have similar degree of functional constraints. It is correct to say that the 50% changes between two substrains of yeasts are neutral changes for both substrains. So the modern evolution theory is a theory of microevolution or a theory of population genetics, relevant only to divergence of similar or identical organisms. It does not describe macroevolution because it fails to take into account the obvious fact that functional constraints on mutations differ tremendously in different kinds of organisms.

The idea of functional constraints is not new and is widely accepted. But most people regard gene function as merely that which can be assayed in a test tube and assume the maximum limit on mutation/distance has not been reached during evolution. But test tube function is merely a bare bone function. The functional role of a gene in a test tube is much less than that in a live cell, which is still less than that in an organism with 100 distinct cells types, which is still less than that in an organism with 10000 cells types, and so on it goes …Two variants of a gene may not show any difference in a test tube or in a single cell organism but may be life and death/disease in a human being.

My paper here on the fossils (now published in Riv Biol. 2008, 101: 93-108, after more than a year of peer review) falsified the clock interpretation of the genetic equidistance result and the fundamental notion of the modern evolution theory that genetic distance has always increased with time in the past history of life on Earth. Every biologist today would predict that the genetic distance between ancient fossil organisms should be smaller than their extant descendants. But my finding shows the opposite: the genetic distance between ancient birds (dinosaurs) and ancient mammals (mastodons) is greater than extant birds and mammals. I do not see how this information did not add anything useful to human knowledge. It is easily among the two most earth shaking results in evolution since Darwin. The first is the genetic equidistance result that is completely unexpected from classical new-Darwinism. The incorrect interpretation of this result known as the molecular clock has created the modern evolution theory. The fossil result of mine is completely unexpected from the modern evolution theory and would not have been earth shaking if we had in the beginning understood the first result correctly.

My result certainly did not add anything useful to the modern evolution theory but that is to be desired if that theory is not completely correct. We often infer what happened in the past but the fossil can tell us directly the story in the past. It is beyond me if someone cannot see the value of fossil sequence in testing the claims of the modern evolution theory or any evolution theory for that matter. Does the theory predict what the fossil data should be like? If it does, then let’s test it as I did. If it does not or only give vague predictions or predicts everything, then it is useless, non-scientific, and irrelevant.

Your comments indicate that you have not read my paper carefully, since my paper has addressed your point. I quote from my paper: “while my analysis showed that humans and chimpanzees are equidistant to gorillas, it also showed that ancient Neanderthals are significantly more distant to gorillas than chimpanzees are. It suggests that the difference between Neanderthals and chimpanzees in their distance to gorillas is outside the variation range that is allowed by the genetic equidistance result. By the same standard that is used to justify the conclusion that humans share the same distance to gorillas as chimpanzees, Neanderthals clearly do not share the same distance to gorillas as chimpanzees.”

My finding reports Neanderthals as a violation of the genetic equidistance result. It has nothing to do variable clock since the equidistance result holds regardless of variable clocks or not. See my paper posted here ‘The genetic equidistance result is independent of mutation rates.’ None of the results showing variable clocks violated the genetic equidistance result. The equidistance result is not an outcome of a constant clock and is independent of clock variations. Most of the variable clock results interpreted minor deviations from exact equidistance as being meaningful when they are in fact not. In short, the genetic equidistance result has no known violations among all extant organisms but is violated by all three ancient fossils for which we have sequence data. Such violations cannot be explained by variable molecular clock since the equidistance result has nothing to do with clock rate variations. I have shown that genes or organisms with variable clocks still do not violate the equidistance result.

The recently published Cell paper on the full mitochondrial genome of Neanderthal is completely consistent with my paper and provided more and independent evidence for my conclusion (Cell 134, 416-426, 2008). The Cell paper states:”A striking observation from the analysis of the 13 protein-coding genes in the mtDNA is that the ratio of nonsynnonymous to synonymous evolutionary rates is significantly higher on the Neandertal lineage.” In other words, at the amino acid level, Neandertals are SIGNIFICANTLY more distant to the outgroup chimpanzees than humans are. Of 7 informative proteins, 6 showed Neanderthal to be more distant to chimpanzees than humans are while 1 showed less. I of course have already found this by studying the hyper variable regions, which has control functions and therefore are non-neutral sequences.

For neutral or synonymous sequences the Cell paper did not find significant difference in distance to chimpanzees for the neandethals relative to humans. So the clear difference between synonymous and non-synonymous sites is unexpected from the molecular clock hypothesis. The hypothesis is supposed to say that most amino acid changes are neutral.

Your comment suggests that you want to explain away my finding by the non-testable ad hoc speculation of a faster mutation rate for the Neanderthals. Unfortunately for you, such speculation has been proven wrong by the Cell paper, which shows that neutral/synonymous mutations have normal mutation rate in the Neanderthals while non-synonymous mutation rate seem to be higher. It is not possible to imagine that a fast mutation rate should only apply to non-synonymous sites but not to synonymous sites. Probably for this reason, the authors of the Cell paper did not invoke the variable clock idea to explain their data.

Instead, they interpreted their ‘striking observation’ by another equally wild and speculative ad hoc idea of less purifying selection due to an imagined small population size for the Neandertals than modern humans. Such idea is not testable and hence not scientific. It can never be proven correct. It is extremely strange that population size becomes an issue only when we are dealing with fossils. All hundreds of extant primate species are equidistant to an outgroup such as rodents. To explain this equidistance by the clock idea, we would have to assume that all these primate species have similar population size as well as similar mutation rates. If so, then it would be extremely rare or strange for a primate species to have significantly different population size from others. It is even more strange then that the only exception should turn out to be the only ancient fossil primate for which we have sequence data. The truth is that population size has nothing to do with the equidistance results. It is pretty safe to say that different species have very different population size just like having very different mutation rates. If the equidistance result is not all related to population size or mutation rates, then of course any violations of the result, such as findings in my paper and the Cell paper, cannot be related to population size or mutation rates.

When one does not have the truth, one simply cannot tell a coherent story. Using an ad hoc idea to explain away one un-expected result would inevitably be contradicted by another fact (s). The idea of less purifying selection for the Neanderthals has another major contradiction. If the equidistance result is an outcome of constant clock, then it cannot be related to natural selection. The constant clock idea requires that most mutations are neutral. So, if the equidistance result is a consequence of neutral mutations, it simply cannot be a consequence of purifying selections. Therefore, any violations of the equidistance result simply cannot be explained by any selection schemes if one accepts the molecular clock hypothesis.

The genetic diversity of Neanderthals in mtDNA is known to be similar or slightly greater than modern humans or Europeans. This is against the idea of a small population size since small population size should have lower genetic diversity. The authors of the Cell paper also wrote some of those early papers on Neanderthals genetic diversity.

Krings, M., Capelli, C., Tschentscher, F., Geisert, H., Meyer, S., von Haeseler, A., Grossschmidt, K., Possnert, G., Paunovic, M., and Paabo, S., 2000. A view of Neandertal genetic diversity. Nat Genet 26, 144-6.

Orlando, L., Darlu, P., Toussaint, M., Bonjean, D., Otte, M., and Hanni, C., 2006. Revisiting Neandertal diversity with a 100,000 year old mtDNA sequence. Curr Biol 16, R400-2.

It is very self-serving for them to ignore these papers that are damaging to their speculation of a small population size. They act as if those papers do not exist by not citing these relevant papers and by stating the following: ‘Future work will reveal if a small effective population size is compatible with the extent of nucleotide diversity seen in the Neandertal nuclear genome.’ Well, existing data have already shown at least for mtDNA that the idea of small population size is invalid.

No theory in the molecular evolution field right now can explain more than half of all data. The constant clock idea explains away the equidistance result but is contradicted by all other data showing variable clock. The variable clock idea explains away a lot of data but cannot explain the most fundamental result, the genetic equidistance result. One should always aim at a theory that explains everything and is contradicted by nothing. The new MGD hypothesis is the only theory that can explain every relevant facts of evolution in a consistent and coherent fashion. It does not invoke any ad hoc speculations or just so stories. It makes numerous precise predictions which have all been confirmed. There is no known factual contradiction. Best of all, the hypothesis is self-evident and stands valid based on logic alone. By contrast, the constant clock idea is against all reason and logic and is openly acknowledged by experts as ‘unthinkable’.

For a true explanation of the equidistance result and the fossil violations of the result, see my paper on the MGD hypothesis posted here ‘Inverse relationship between genetic diversity and epigenetic complexity.’ The main idea of this paper has now been published as a peer reviewed book chapter: Huang, S. (2008) Histone methylation and the initiation of cancer. Cancer Epigenetics, Ed. Tollefsbol, T., CRC Books.

No comments: