More Recent Comments

Monday, January 02, 2017

You MUST read this paper if you are interested in evolution

A reader alerted me to a paper that was just published in BMC Biology.1 The author is Eugene Koonin. He makes the case for neutral evolution (random genetic drift) and against adaptationism. You may not agree with his take on evolutionary theory but you better be aware of it if you claim to be knowledgeable about evolution.

Koonin, E.V. (2016) Splendor and misery of adaptation, or the importance of neutral null for understanding evolution. BMC biology, 14:114. [doi: 10.1186/s12915-016-0338-2]
The study of any biological features, including genomic sequences, typically revolves around the question: what is this for? However, population genetic theory, combined with the data of comparative genomics, clearly indicates that such a “pan-adaptationist” approach is a fallacy. The proper question is: how has this sequence evolved? And the proper null hypothesis posits that it is a result of neutral evolution: that is, it survives by sheer chance provided that it is not deleterious enough to be efficiently purged by purifying selection. To claim adaptation, the neutral null has to be falsified. The adaptationist fallacy can be costly, inducing biologists to relentlessly seek function where there is none.
While explaining the limits of natural selection in smaller populations, Koonin makes the point that "Nothing in evolution makes sense except in the light of population genetics" (Lynch, 2006). He points out that,
... although most biologists do not pay much attention to population genetic theory, the time seems to have come for this to change because, with advances in functional genomics, such theory becomes directly relevant for many directions of experimental research.
This is important. Evolutionary biologists cannot continue to ignore the fundamentals of evolutionary theory developed by population geneticists over the past century. As we saw in London last November, there are too many so-called "prominent" evolutionary biologists who are ignorant of modern evolutionary theory and the null hypothesis. Ironically, these same biologists are the ones leading the charge for changes in evolutionary theory!

Koonin also makes an interesting point about species with historically small population sizes. These species accumulate a lot of junk in their genomes because natural selection is powerless to remove it. This junk DNA provides a pool of potential exaptations that usually lead to an increase in complexity (constructive neutral evolution). Thus, paradoxically, a weakening of natural selection often leads to more complexity.

A typical example involves gene duplications. In prokaryotes, with large population sizes, the duplicated gene is often eliminated by negative selection because of the slightly deleterious effect of having an extra gene in the genome. In mammals, however, that deleterious effect doesn't meet the threshold for selection and the gene copies persist in the genome. Over time they may acquire independent functions by specialization and now both genes are essential. Thus, the organism has become more complex because it now needs a gene family whereas only one gene was required in species with larger population sizes.

You MUST read this paper in order to understand the real debate within evolutionary theory. It's a very good summary of the point of view I've been trying to communicate for a very long time. I'm just a messenger but Koonin is one of the players.

1. I don't have permission to reveal their identity.


John Farrell said...

His Logic of Chance book is excellent.

Unknown said...

A new trait can be very hard to recognize, especially if you don't know what you are looking for.
Compared to humans, the blind watchmaker has a 20/20 vision.

S Johnson said...

Jerry Coyne has recently and helpfully cited literature that refutes criticism of the notion that "most important evolutionary
changes at the level of the visible
phenotype, as revealed by paleontological
and systematic studies, have resulted from
natural selection acting on variation within
populations. This variation is ultimately
due to mutations that arise at random
with respect to the direction of selection.
A role for evolutionary forces other than
selection and mutation, such as random
genetic drift, is of course recognized by
neo-Darwinists, but selection is regarded
as the main guiding force of phenotypic

The professional view of Gould is pretty much the same view of other critics of adaptationism (via natural selection, which includes sexual selection) so far as I can tell. In that view Gould was a self serving con artist, who also falsified date for political reasons in The Mismeasure of Man. It's hard for a layman to be absolutely certain of such things as prevailing professional opinion, but it seems to me that Gould is still regarded as the closest thing to a serious critic of adaptationism, given the support from paleontology. Thus the importance of this timely refutation of the strongest case against adaptationism.

Personally, I think natural selection is hugely important, possibly the most important practical factor, because it staboilizes phenotype. But no doubt that's a deadly combination of ignorance and arrogance.

Dave Carlson said...

I'm sure that the paper is worth reading (I've skimmed much of it already), but good lord that 3rd figure is... distinctly lacking in certain pleasing aesthetic qualities.

Larry Moran said...

In that view Gould was a self serving con artist, who also falsified data for political reasons in The Mismeasure of Man.

The book is about racism. Gould was attacking the idea of genetic determinism and the concept that some races are more intelligent than others. After the book was published the racists defended themselves by criticizing Gould's methodology and defending their own views.

It's a complicated subject and there are plenty of errors on both sides. Nevertheless, Gould's main point is correct and he did NOT falsify data in either the first edition (1981) of the second (1996).

Here's a summary of the controversy: The Mismeasure of Man

Donald Forsdyke said...


Following a multidisciplinary study of milk production at a dairy farm, a physicist returned to explain the result to the farmer. Drawing a circle she began: "Assume the cow is a sphere … ." This insider math joke may explain Koonin’s puzzlement that "most biologists do not pay much attention to population genetic theory" (1).

The bold statement that "nothing in evolution makes sense except in the light of population genetics," cannot be accepted by biologists when evolution is portrayed in terms of just two variables, a "core theory" that involves "an interplay of selection and random drift." While mathematical biologists might find it "counterintuitive" that "the last common eukaryotic ancestor had an intron density close to that in extant animals," this is not necessarily so for their less mathematical counterparts. They are not so readily inclined to believe that an intron "is apparently there just because it can be" (2).

While expediently adopting "null models" to make the maths easier, population geneticists are not "refuted by a new theoretical development." They have long been refuted by old theoretical developments, as illustrated by the early twentieth century clash between the Mendelians and the Biometricians (3). It is true that by fiddling with "selection coefficient values" and accepting that "streamlining is still likely to efficiently purge true functionless sequences," the null models can closer approximate reality, but a host of further variables – obvious to many biologists – still await the acknowledgement of our modern Biometricians.

1.Krauss LM (1994) Fear of Physics: A Guide for the Perplexed. Jonathan Cape, London.

2.Forsdyke DR (2013) Introns First. Biological Theory 7, 196-203.

3.Cock AG, Forsdyke DR (2008) "Treasure Your Exceptions." The Science and Life of William Bateson. Springer, New York.

S Johnson said...

For what it's worth, I agree that the case for lying made against Gould fails. But then, since I also believe that the case against punctuated equilibrium as a phenomenon that demonstrated natural selection for adaptations is not the main driver of phenotypic change (what naive lay people would call multiplication of species) wasn't really refuted in 1982. So I guess my opinion is of negative value.

Diogenes said...

Certainly Gould did not falsify any data. Rather, he used statistical methods to detect bias in the work of a long-dead scientist who measured the sizes of skulls which are now rarely accessible (the actual skulls are now somewhere in the archives of the Academy of Natural Sciences of Philadelphia, I believe.) Based on this Gould accused Samuel Morton, a scientific racist, of biasing his data set against non-whites. The question was whether Gould was using the proper or improper statistical tests.

The Intelligent Design advocates, who have always wanted to take down Gould, leaped to the defense of Samuel Morton, indignantly defending the long-dead racist as an aggrieved innocent victim of the recently-dead Gould. Gould concluded that the racist was *unconsciously biased* against non-white skulls.

At worst, Gould, who was not a professional statistician, used the wrong statistical test. This is a common argument in science and has nothing to do with faking data.

Note Gould did not toss out accusations of deliberate fraud against the long-dead Morton. However, creationists and ID advocates have tossed out casual accusations of scientific fraud against the more recently-dead Gould, whom they need to destroy. Neither is alive to defend themselves.

(For what it's worth, I'll point out that what Gould is accused of is nothing compared to the incompetence and fraud that goes on among creationists and ID advocates, the people who brought us the Paluxy River man-prints and the laughable claim that the human and chimp genomes are only 70% similar.)

If it were proven that Gould's recent critics themselves used the wrong statistical test, or even faked their data, they will never be accused of scientific fraud, without accountability, as casually as they have accused Gould of same. Why do you think that is?

Larry Moran said...

I also believe that the case against punctuated equilibrium as a phenomenon that demonstrated natural selection for adaptations is not the main driver of phenotypic change (what naive lay people would call multiplication of species) wasn't really refuted in 1982.

I can't parse this sentence. It suggests you don't understand punctuated equilibria. Is that what you meant?

Marleen said...

I wonder if it is possible when we talk about and evaluate traits as adaptations to consider the genotype, the genes and introns. Isn’t it true that adaptions regard only the phenotype? Can we consider introns and their evolution as traits that are yes or no under selection? What do you think of Koonin dedicating a whole paragraph to the evolution of introns in an article with adaptation as main subject?

When E. Koonin writes the following:

"Introns once again present a perfect example. Because introns cannot be efficiently eliminated by selection, eukaryotes have evolved, first, the highly efficient and precise splicing machinery, and second, multiple lines of damage control such as nonsense-mediated decay, which destroys aberrant transcripts containing premature stop codons. In a more speculative vein, the nucleus itself may have evolved as a damage-control device that prevents the exit of unprocessed transcript to the cytoplasm”.

Since we do not know why the ‘highly efficient and precise splicing machinery’ and the nucleus evolved, it seems to me an adaptionist view to assign them this particular role and function,

I agree very much with the comments of S Johnson

Unknown said...

Some thoughts on the article:
1) "exaptation is a new term introduced by Gould and Vrba to denote gain or switch of function during evolution". Not that new anymore, as would be made clear by citing Gould and Vrba (1982) "Exaption - a missing term in the science of form", Paleobiolgy 8,4-15.
2) "Seemingly not—in particular because clear-cut examples of spandrels are notoriously difficult to come up with. Nevertheless, the essential message of Gould and Lewontin, that telling just-so stories is not the way to explain biology, stands as true and pertinent as ever in the post-genomic era. Let us explore the reasons for this, which could actually be simpler and more fundamental than those envisaged by Gould and Lewontin."
If clear-cut examples of spandrels are hard to find (and I'm not convinced they actually are that hard to find), it would not really diminish Gould and Lewontins main argument, which is that there are often more traits than parameters for a given morphology (for instance a spherical organism can be described with one paramter: the radius of the sphere. On the other hand we could look at a lot of traits - diameter, surface area, volume, Ratio between Area and Volume, etc. all of which might be biologically relevant). The essential message of the spandrels paper is about spandrels. Why spandrels have relevance in genomics eludes me (more on this later).
3) "If |N_e s| >> 1, the mutation will be deterministically eliminated or fixed by selection, depending on the sign of s."
This is incorrect in a lot of cases. For s=.05 and N_e=10^200 I get a probability of fixation of 0.095, although the condition given is certainly met and Haldanes approximation (which Larry keeps using) gives us 0.1. Even s=1 makes it just go up to 86%. If you make bold claims of determinism for large |N_e s|, you are not making a case for the importance of drift, but against it.
4) Fig 3, is not only ugly as note by a previous poster, but it also sharply divides the N_e-s phase space into two "regimes" one "selection dominated" and one "drift dominated". While my basic objections to treating drift and selection as separate are already known, here I would mostly like to point out that this sharp divide is not even backed by Koonins own argument, which discusses |N_e s| < 1 and |N_e s| >> 1. The graph here implies that the "determinism" proposed for large |N_e s| already holds for |N_e s|>1 (making a "larger than" out of a "much larger than").

Unknown said...

5) "These routes of exaptation that appear to be central to eukaryotic evolution notably deviate from Gould’s and Lewontin’s original spandrel concept".
They do not just notably deviate, they are completely different. It's hard to argue that individual nucleotides are not independent "traits" and therefore it is also hard to argue that there a sequence level spandrels. Koonin correctly uses exaption here, but somehow seems to argue that spandrels are a prerequisite for exaption. But that is not how Gould and Vrba define the term (in fact they only cite the spandrels paper in passing and their examples for exaption are mostly non-spandrels). Under 1) I argued that it would have been helpful to cite the Gould and Vrba paper, here I would go further and suggest that it might have helped to read it.
6)"The population genetic perspective calls for a change of the null model of evolution, from an unqualified adaptive one to one informed by population genetic theory, as I have argued elsewhere."
Wait, is there any literature in which anything but a neutral model is used as a null? I'm pretty sure that neutral nulls were the norm even before Koonin presented his argument in 2004.
7) "Null models are standard in physics but apparently not in biology."
Citation needed.
8) "However, if biology is to evolve into a “hard” science, with a solid theoretical core, it must be based on null models, no other path is known."
It's more than mildly infuriating to read this, since we have both a solid theoretical core in the form of population genetics and there are literally thousands of papers using neutral null models. Heck, most of the google groups for software packages I'm using is people asking about how to set up their statistical tests vs. neutral nulls.

Dave Carlson said...

I also though it was odd that Koonin wrote this as if neutral null models are not already incredibly common. Perhaps his intended audience is largely people who don't do molecular evolutionary/pop gen research? Still the claim that null models are not standard in biology seems suspect, at best.

Larry Moran said...

The ENCODE fiasco is just one of many where the null hypothezis was function/adaptation. Very few evolutionary biologists are familiar with drift and Neutral Theory. Koonin is correct.

Dave Carlson said...

"Very few evolutionary biologists are familiar with drift and Neutral Theory."

Is there a word missing from this sentence? As written, it's certainly incorrect.

txpiper said...

Only one of Dr. Moran's posts above is functional.

Larry Moran said...

@Dave Carlson

I was at s Royal Society meeting last month. It was full of evolutionary biologists who think that natural selection is the only game in town. I'm sure that most evolutionary biologists have heard of drift and Neutral Theory but that doesn't mean they have incorporated the concepts into their way of thinking.

Next week I'll get a chance to quiz 40 students on their knowledge of evolution based on their first and second year courses at the University of Toronto. In the past, most of them cannot recall hearing about drift and Neutral Theory because their evolutionary biology professors didn't think they were important enough to emphasize.

Mikkel Rumraket Rasmussen said...

Your point about the spliceosomal complex makes little sense ot me. It's pretty clear that introns, if left alone, would be highly deleterious (particularly now when they are ubiquitous in eukaryotic genes). That would make the splicing machinery a bona fide adaptation to their (initially, low frequency) presence once upon a time.

The part about the role of the nucleus is more speculative, but he of course says that himself. Whatever you may think about the merits of such a take on the nucleus as a whole, the deleterious effects of introns specifically, if not removed, and the function of the spliceosomal complex as a defense-mechanism seem to me pretty much indisputable.
After all, it really is a fact that if introns aren't spliced out, most eukaryotic protein-coding genes would become nonfunctional. I light of this, if anything is obvious, the evolution of splicing machinery as an adaptation to their presence and proliferation, is.

Marleen said...

I agree with you that introns are and may have been deleterious now and at the onset of evolution of eukaryotes. It is possible however to speculate that there has been a co-evolution between the appearance of introns in genes and the formation of a splicing machinery. In that case I cannot see any adaptation to a deleterious event but simply the co-evolution of two properties together; introns and splicing machines.

Koonin indeed wrote that it is pure speculation to suppose the nucleus is formed to create a separate space for non-processed transcripts. What surprises me is that he 1. brings up this very speculation so easily in this case, and 2. that he assigns a function to the nucleus which means he considers the nucleus an adaptation without any proof of that, neglecting his own claim of neutral null.

Marleen said...

@Laurence Moran

If one looks up "genetic drift" or "random drift" in one of the handbooks of evolution (p.e. Barton’s “Evolution” 2007) there is a substantial chapter (chapter 15) that treats genetic drift. Therefore, although I was not at the meeting, I cannot imagine that evolutionary biologists do not consider it a serious factor in evolution. Apparently the subject is part of the curriculum and if these evolutionary biologists are teachers they certainly do teach genetic drift. If they consider natural selection all the same as the major drive of evolution is not only their free choice, but also an understandable choice, because this is one of the major disputes in the biologists community.

Dave Carlson said...

I wasn't at that meeting, but I know and interact with a lot of evolutionary biologists on a daily basis. Every single one of them is familiar with drift and neutral theory. Whether or not they all incorporate those concepts into their thinking in a way that would satisfy you is unclear. But that wasn't the original claim you made.

In addition, if you read any number of molecular evolution/evolutionary genetics papers, you'll see that neutrality is often assumed or used as a null model. Hence my suggestion that Koonin's paper is probably focused on people working outside of those disciplines (e.g., functional genomics researchers).

Larry Moran said...

@Dave Carlson

I agree that many papers in molecular evolution and evolutionary genetics will consider neutrality as the null model and they demonstrate a working knowledge of drift. However, most evolutionary biologists are not working in those areas.

Even in those areas, working knowledge of drift and Neutral Theory isn't as widespread as you think. A few years ago, I interviewed graduate students and post-docs at an SMBE meeting. I asked them to describe evolution and about one third defined it as natural selection. On further questioning it turns out they knew something about drift but they had obviously not incorporated it into their thinking. Adaptationist thinking is still rampant in the discipline. Koonin is correct.

Larry Moran said...

The original introns are thought to be Type II self-splicing introns derived from the primitive mitochondrion. They were only slightly deleterious because they excised themselves from the primary transcripts. Because they persisted in the genome, it led over time to the gradual loss of self-splicing activity and evolution of trans factors to assist in splicing.

Thus, the evolution of a complex spliceosome is due to the fact that the original introns could not be eliminated by negative selection. In other words, the weakness of selection causes complexity.

Unknown said...

The ENCODE fiasco is just one of many where the null hypothezis was function/adaptation. Very few evolutionary biologists are familiar with drift and Neutral Theory. Koonin is correct.

The ENCODE fiasco was mainly due to their redefinition of function. I'm not sure Koonin understands what he is talking about - he is severely underestimating the role of drift in his "selection dominated regime" and he also conflates the concepts of spandrels and exaption (where the only relevant connection is that both were co-coined by Gould).

Anonymous said...

I'm sure Simon is correct that those who model evolution mathematically use drift as the null hypothesis, but that is a small subset of those who study evolution in some way. Many biologists (and more chemists, etc.) think of evolution only or mainly in terms of selection. I know my own transition from "yes, drift happens but . . ." to an appreciation of its pervasiveness and role in development of new traits was slow. So anyway, there is reason for Koonin to write his article.

Veiko Krauß said...

The problem might be that many biologists may be able to correctly define drift if questioned, but do not accordingly integrate this and other evolutionary factors into their overall model of evolution. Even an eminent evolutionary biologist as Andreas Wagner do not like to call his networks of mutations with the same phenotype neutral networks. Why not? I suppose that ideological adaptionist preoccupations are so widespread and influential that they may often suppress seemingly unavoidable scientific conclusions.

rich lawler said...

The claim "the null model needs to be falsified" is pretty easy to make, but can be difficult in practice to achieve, particularly if the null model itself is full of assumptions. Null models have long been a part of evolutionary ecology, with null models proposed for phenotypic evolution, life history evolution, and species distributions/area relationships. For example, Mike Lynch (and Hill) developed a model of neutral phenotypic evolution (based on mutational input and drift) but this null model is pretty difficult to falsify even in cases when we know the trait in question is likely under selection (e.g., increases in human brain size); too many assumptions go into it. Similarly, there is a new set of models put forth by Hal Caswell (and independently by Tuljapurkar/Steiner) on neutral life history evolution. However, these models make assumptions about how organisms move through their life cycle (a 1st order Markov assumption--current state depends only on prior state), which is, for some, too stringent of a condition to impose on organisms, and too stringent to suggest the model can serve as a null model. There is also a long history of null models in ecology (as reviewed in the book "Null Models in Ecology").

In all cases, the null model must be mechanistically well-justified and the assumptions clearly laid out. Some folks would argue that in the cases above, the assumptions don't justify the model as being labled a "null" model just because the model invokes drift, neutrality, or some other stochastic-generating component. Null models themselves deserve scrutiny. In the case of molecular evolution we are probably on surer footing, since the assumptions going into the neutral theory are better justified and mechanistically sound (though John Gillespie would disagree with this).

Graham Jones said...


On point 3, I think Koonin is looking at the equilibrium distribution, not the fate of a single new mutation. By 'equilibrium distribution' I mean Pr(i out of N are a particular allele) for 0 <= i <= N, after a long time.

Unknown said...

@Graham: In that case his choice of words is extremely misleading, since he states that fixation or elimination of a new mutation becomes deteministic, which at least strongly implies he is talking about origination-fixation models there. There's also the issue that even if a stationary distribution exists, it's unlikely to be reached in any reasonable time. If we discount mutations, we do not even have an irreducible MC, which means that there is no single recurrence class and thus no stationary distribution.

Graham Jones said...

Simon, I am not defending what he said, only trying to work out what he was trying to say. Sometimes authors say one thing in words, and another in maths, and leave it as a puzzle for the reader to decide whether to work forwards from the words to correct the maths, or backwards from the maths to correct the words. (I think you know this all too well, but others may be reading.)

I wouldn't discount mutations. That would be weird.

As to whether there has been time for stationarity to be reached, it will be different for different cases, and I think it's an interesting set of questions that are too complicated to answer here.

S Johnson said...

Deleted response with typo! Corrected version follows.

The sentence refusing to be parsed was referring to Coyne's attack on Gould, and his citation of a refutation.

Coyne wrote: "Steve Gould and Niles Eldredge, with their theory of punctuated equilibrium proposed in the Seventies, essentially made the extreme non-Darwinian claim that big evolutionary changes happens when small populations somehow lose their 'genetic equilibrium' (how wasn’t specified), and, further, that species selection was responsible for trends in the fossil record as well as adaptations themselves." (Original at

I had thought what punctuated equilibria, as demonstrated in the fossil record, was the stabilizing role of natural selection as the general rule, which is why natural selection as the main driver of optimal adaptations seemed to me to be, well, not quite obvious. But since I do not understand it after all, as you tell me, then the refutation made so long ago must be entirely correct. That is at

Multiplication of species is what most lay people are primarily interested in when thinking about evolution. Most accept on authority natural selection has pervasively perfected traits.

As to personal opinions of Coyne, since I do not understand punctuated equilibrium I must agree with you and Coyne. "I [Coyne] wanted to add one additional reason why evolution is liable to such critiques: scientists are always out to make a name for themselves, as our currency of achievement is not money but reputation. You don’t get well known by just adding another brick to the evolutionary edifice, but you can do by pushing the wall over. That’s how Steve Gould made his name, flawed as his theories were. With all its messiness, poorly understood phenomena (what are females choosing during sexual selection?) and historical contingency, you can always assemble a list of phenomena that you can claim show severe deficiencies in evolution. But as with theological arguments, a lack of explanation doesn’t mean that we have to resort to drastic conclusions."

Lastly the date 1982 was not a reference to any papers by Gould, Eldredge or Lewontin, but to the conclusive refutation so unjustly overlooked by Gould's fans for these many decades. I thought I was being ironic but plainly the jokes on me.

Steve said...

It really doesnt matter if one leans toward neutral theory, adaptation, natural selection, yada yada.

any theory of evolution can only ride on the coattails of design.

I have not come across one person (including the IDiot slayer Moran) who can logically articulate how the first simple organism could possibly evolve considering any of the above explanations without design at its core.

Will Moran be the first?

Now that Larry will retire soon, maybe he'll have the time to take a crack at a narrative that describes logically sound scenarios that can explain the complexification of organism without any bells and whistles to work with.

Anyway, dont get me wrong. Im fine with design deniers slugging it out to see who becomes king of the 1%ers.

Faizal Ali said...

Hey there, Steve. Would you care to try answer a question I have posed to several other of your fellow proponents of the idea that life was created by a "designer"? So far, no one has even given it a try.

Every "designer" we know of is a living being. Therefore, all available evidence suggests that life has to exist before a "designer" can. So how is it possible for life to have been created by a "designer", if designers only arise after life already has?

I look forward to your reply.

Unknown said...

I believe this is convincing argument for the ‘neutral’ hypothesis as the default assumption:

If one assumes the ‘drift’ hypothesis, then one can test to see if there is an actual advantage to the sequence or not. If one can find an unequivocal advantage, then one can claim ‘selection’.

If one assumes there must be an advantage to the sequence (selection hypothesis), then one can never falsify this hypothesis (there could always be some other ‘advantage’ we haven’t thought of yet).

Therefore in order to have a testable, falsifiable hypothesis, we should stick with the ‘neutral’ hypothesis regarding genetic sequences.

TheOtherJim said...

Trying to play the moderator here. IT is true that those studying molecular evolution typically involve NT in their work, as this has always been the bedrock of the field. But in the more Ecology/Animal Evolution world, I would propose you will have a hard time finding examples of neutral mechanisms being the null. And fun stories of mammalian “X probably evolved for Y” are the non-specialists day-to-day familiarity with Evolutionary Biology.

And underlying the ENCODE phenomenon - we have an entire discipline of "Molecular Biologists" out there now from various disciples who have had no real formal training in modern evolutionary thinking. They remember Darwin, Origin of Species and the "Survival of the Fittest" from intro biology classes and move forward as though their understanding is complete. I see this daily in my field, and in the intro lectures I give in a course here. The students are stunned at the implications of the classic simulations like the Driftworm simulation . They have just never thought about it, having only heard the “why a zebra has stipes” stories.

Setting aside the semantics of the ENCODE debate, this underlying assumption they have the whole evolution thing worked out leads them to the sorts of claims you see in the "headline" ENCODE papers. And I think this is unique to evolution. I would never dare to write a about how my molecular-genetics work solves great mysteries of Developmental Biology. But somehow, everyone with a half-hour intro to evolution under their belt is quite willing to make grand proclamations about their RNA-Seq data and the implication of this for Evolutionary Theory.

Arlin said...

@Marleen, you may be right that students today learning about drift, but Larry went to a conference of *old people*, the sort who attend conferences about the state of evolutionary theory. If you are over 50, then when you went to university, "molecular evolution" (including the neutral theory) was a specialty topic.

Arlin said...

@Simon, yes Koonin is mistaken about "deterministically eliminated or fixed by selection". The statement should be "deterministically eliminated by selection or stochastically fixed (with p ~ 2s)". There is a widespread misconception that the probability of fixation relates in some strong way to population size. Instead it is almost entirely a matter of how close the number of alleles is to 0, not how far it is from 2N.

Pauline said...

Here is a "no one has even given it a try" reply lutesuite: An analogy: "Every artist we know who paints pictures is a painter. So who painted the Bayeux tapestry?" Well the premise that the Bayeux tapestry was "painted" is wrong. So the question cannot be answered.
So your question presupposes that every 'designer' is a living being. God is not a "living being" in the set of living beings that you describe. Christians believe that God is outside of creation.

A common question from Atheists is "If God create the world, then who created God?" This presupposes God was "created", but christians believe that God existed before time began and was not created (ignore the fact that we as humans cannot understand this - why should we?)
Now let's turn this on evolutionists - if all theories work on existing life (eg natural selection, survival of the fittest) then how did the first life 'evolve' from non life?

I look forward to your reply

Faizal Ali said...

Very good, Pauline. You've explained why the premise of the creationist "chicken and egg" arguments often made here is false. I'm glad you seem to understand where I'm coming from. Maybe you could try explain this to your creationist buddies like Steve and txpiper, so they could stop embarrassing themselves and setting such a bad examples for religious believers.

And, of course, since you're so clever you clearly only intended your question to "evolutionists" as an illustrative example of the kind of moronic questions creationists less enlightened than yourself are prone to asking, and don't really need anyone else to explain to you why it's such a stupid question.


Misconceptions in evolution theories said...

"Since we do not know why the ‘highly efficient and precise splicing machinery’ and the nucleus evolved, it seems to me an adaptionist view to assign them this particular role and function,"

"It is possible however to speculate that there has been a co-evolution between the appearance of introns in genes and the formation of a splicing machinery."

That is true. But it is also possible to speculate that the splicing machinery is older than the exons. We do not know the origins of introns or the splicing machinery, and we do not know the origin of the nucleus. Thereby we cannot know much about their early evolution. But we can know about their more recent evolution. Regarding their origin, we have to also consider the possibility that introns were the original, and exons the novelty. In that case the splice machinery was needed to pick out the introns and to use them as different sorts of RNA. The spiced waste was only useful as mRNA after invention of the translation machinery.

Regarding the nucleus, it may have been the original, and cytosol may have been the novelty. It was only needed as a place for the translated RNA, i.e. proteins to operate.

Misconceptions in evolution theories said...

"If one looks up "genetic drift" or "random drift" in one of the handbooks of evolution (p.e. Barton’s “Evolution” 2007) there is a substantial chapter (chapter 15) that treats genetic drift."

That is very good. But why are so few authors writing about it or using it in their research? My impression is that in nearly all microbiological or molecular research articles, when it is discussed how their results relate to evolution, then it is "committee evolution" (ref. Mayr, Dobzhansky et al) based on selection it is referred to. Who except Masatoshi Nei, Arlin Stoltzfus and Koonin are today promoting the neutral theory? There was not too many in the twentieth century either. Motoo Kimura opened up my eyes for what evolution really is with his neutral theory in 1968. Before that I thought evolution was all about selection.

"I agree very much with the comments of S Johnson"

That is interesting. Can you tell me what really is the essence of his comments?

Dave Carlson said...

@Jarle Kotsbak
"That is very good. But why are so few authors writing about it or using it in their research? My impression is that in nearly all microbiological or molecular research articles, when it is discussed how their results relate to evolution, then it is "committee evolution" (ref. Mayr, Dobzhansky et al) based on selection it is referred to. Who except Masatoshi Nei, Arlin Stoltzfus and Koonin are today promoting the neutral theory? "

Molecular Biology and Evolution is the most highly cited journal in evolutionary biology (I think this is largely due to the publication of papers describing certain widely used bioinformatics programs, but let's leave that aside for the moment). If you do a keyword search for neutral theory or genetic drift in that journal, you will see that the topics come up extremely frequently in research articles.

Marleen said...

@Jarle Kotsbak

Strictly spoken your are right: since we know nothing about the origin of introns and exons we may speculate whatever we want. It seems to me however that the current hypothesis is that eukaryotes evolved from prokaryotes (alfa-proteobacterie was engulfed by an archaeon). Therefore, it seems reasonable to consider exons (the protein-coding part of a gene) as the descendant of the intron-free genes of prokaryotes. Hence the necessity of the evolution of a spliceosome to get rid of this novelty, the introns.

I am not sure that neutral theory and drift should be ‘promoted’. Why should they, if most if not all evolutionary biologist agree that these theories have an important role in evolution.

May I ask you something? How can we tell that a sequence or trait is advantageous, how can we determine the importance of genetic drift with respect to natural selection in the evolution of one particular organism or trait. We do agree that natural selection is important, so what percentage of the evolutionary process was directed by drift and what by selection. Personally I have no idea about how we can determine this.

Marleen said...

How can you tell there is an advantage to (‘having’ or carrying) a sequence. First of all you need to determine what this advantage is, in what it consists. There exists already a problem in my opinion: as soon as you assign a possible advantage to a trait or sequence you claim to know what the advantage of this trait is.

You say that if you assume the ‘drift’ hypothesis and are able to demonstrate there is an ‘advantage’ you may claim selection. Also in this case however you might overlook the possibility of other advantages, hence you may overlook any advantage. Therefore you can never claim the ‘drift’ hypothesis.

Misconceptions in evolution theories said...

"How can we tell that a sequence or trait is advantageous,.."
As it is almost impossible to tell in advance whether or not a sequence is advantageous, there has been a common misunderstanding that it is possible in retrospect to tell this based upon the reproductive success of the possessor. In that way one makes no room for drift. Many researchers like to assume that the population size is so large that they can disregard drift as a cause. They still have a problem, because absolutely neutral changes are also fixed, and they are actually totally dominating in all organisms. So the best answer to your question, seen in retrospect, is that the best and maybe the only way to identify a sequence as advantageous is that it does not change. All base pairs that are not critical change very readily, so a sequence comparison to related and not so related species directly identifies the critical and therefore advantageous parts of the sequence.

I answered your question based on your term "evolution of one particular organism or trait". That indicates you are talking about evolution including creation of novelty and not just adaptation of allele frequencies in a population, which some researchers has also called "evolution". But when you say: "..what percentage of the evolutionary process was directed by drift and what by selection.", then I would say that the question is impossible for evolution. For adaptation it is also difficult to answer, but at least it has a meaning.

I am much more interested in evolution than adaptation, and I am writing a book based on my understanding of how drift and selection works together creating new features and improving old features. My theory, which I have called "contra-Darwinism", explains how both drift and selection are absolutely necessary in creating or improving features.

judmarc said...

because absolutely neutral changes are also fixed, and they are actually totally dominating in all organisms.

All organisms? Including organisms that do have very large effective population sizes, like microorganisms?

Misconceptions in evolution theories said...

"All organisms? Including organisms that do have very large effective population sizes, like microorganisms?"

Yes, all organisms. Neutral changes are independent of selection. Their probability of fixation is independent of population size. Therefore they are used as molecular clocks. The only difference between small and large populations are that fixation takes longer time in large populations. This was shown by Motoo Kimura fifty years ago.

Anonymous said...

All organisms? Including organisms that do have very large effective population sizes, like microorganisms?

Two different concepts causing confusion here, I think. Bacteria with very high effective population sizes loose (nearly) all their junk DNA because it is mildly harmful and in huge populations even very small selection coefficients can have effects. On the other hand, truly neutral changes (e.g. many third bp changes in codons and DNA sequences in bits of DNA needed as spacers only) would continue to change by drift.

judmarc said...

bwilson - So we are talking about neutral *changes* within functional *sequences*?

Misconceptions in evolution theories said...

Yes, neutral changes, or what is usually called mutations. There is a lot of redundancy in most functional sequences. The third bp in codons, that bwilson mentioned is only one example. Such changes are often silent, i.e. produce the same amino aced sequence. But a lot of amino acid sequences could change almost freely with no consequence. There are a few examples of proteins with little room for changes, e.g. the histones. Histones from non-related species therefore have very similar amino acid sequences.

Misconceptions in evolution theories said...

what percentage of the evolutionary process was directed by drift and what by selection.

This question may of cause be answered if you mean what percentage of the genome is conserved due to natural selection compared to all the rest, which changes freely due to genetic drift. In the human genome much less than 1% of the genome is protected from drift by selection. We can say that the constancy is directed by selection and the changes are directed by drift. But whether or not this is a good answer depends what you mean by "evolutionary process". If you mean something active, i.e. changes, then the percentage that is directed by selection is very close to zero.

Misconceptions in evolution theories said...

I made a search in Pubmed on articles in that journal. I found for "neutral theory" 36, for "genetic drift" 104, for "selection" 1716 and for "natural selection" 273. The results are better than with a general search, which gave 446, 4592, 407592 and 10249 respectively. My impression when reading articles is that results are much more often checked against selection than against neutral theory. And the numbers support this impression.

Dave Carlson said...

@Jarle Kotsbak

I would suggest actually searching the journal's website. Searching for "genetic drift" produces results from more than 4,500 articles (link here: A similar search for "neutral theory" produces results from more than 2,500 articles (

In any case, I read articles in MBE regularly, and neutral theory is discussed both explicitly and used implicitly as an assumption or null model on a pretty consistent basis. The results of selection will always, I think, be sexier to a majority of researchers, but that doesn't necessarily mean that neutrality is always neglected.

Unknown said...

I’m suggesting one assumes the drift hypothesis, one would never ‘prove’ the drift hypothesis, but one could disprove the hypothesis. This is a normal situation in science, as Einstein pointed out, (to paraphrase) 'no test could prove general relativity is true, but one could disprove it.'

By ‘advantage’ what I mean is ‘more likely to reproduce’. So if one can prove an organism with a particular sequence is more likely to reproduce than one without, then one has a case for selection.
It would be at that point one would try to find a trait associated with the sequence that would ‘explain’ the selection.

This avoids the ‘just so story’ which we humans seem so good at producing and buying without any evidence beyond the cleverness of the story that explains the ‘just so’.