The Creationists Discover the Nearly Neutral Theory

Billy D (no, not that Billy Dee) is pimping an anti-evolution book by John Sanford. I refuse to link to outright liars, but you can find a link to Bill's blog here. More after the jump.

From what I can gather, the author postulates that the human genome is deteriorating, and this is evidence against Darwinian evolution. This is another argument -- in a long line of many others -- that basically amounts to the same bullshit: natural selection cannot explain the diversity of life on earth.

Interestingly, Michael Lynch used a similar argument (albeit without supernatural forces) in describing the evolution of the eukaryotic genome. Lynch used the nearly neutral theory to show that the complexity of eukaryotic genomes may be the result decreased population size in eukaryotes (relative to prokaryotes). The strength of selection depends on population size (Ns, so selection is stronger in large populations), and eukaryotes have smaller populations than prokaryotes. Therefore, the complex features of the eukaryotic genome were able to evolve due to relaxed selective constraint in smaller populations.

Thinking about it this way, our genome is actually just a degenerated bacterial genome. It's not surprising, then, to see degeneration along the human lineage as our ancestors for the past tens of millions of years have lived in small populations. See, no need to invoke supernatural forces when the real world provides more than enough mechanisms.

(Via The Scientific Indian.)

More like this

I like it! Reduce the effective population size, selective pressure decreases, and the genome 'degenerates' by getting bigger and more complex!

I'm curious how that gets reconciled with population bottlenecks, though. After a bottleneck, don't we typically see reduced genetic variation between individuals in the population? Is there any conflict there, or do we see higher complexity and reduced variation at the same time?

A bottleneck leads to a decrease in genetic variation because there are less individuals in a small population to harbor that variation. Also, deleterious alleles have a greater chance of fixing following a bottleneck because selective constraint is weaker in small populations. A bottleneck usually only lasts for a few generations. What we're talking about in the evolution of the eukaryotic genome is long term "small" population size (I say "small" because we're talking about it relative to prokaryotic populations).

Over at UD, another somewhat related thread concerned "the cost of natural selection" introduced by Haldane and discussed by Kimura when neutral/nearly-neutral theory was first introduced.

To my knowledge, there has never been a consensus answer for this putative dilemma. There seem to be a few minor papers here and there but nothing I could distinguish as the definitive case-closing work. I understand that the acceptance of neutral and nearly-neutral mutations cured the problem of how all those polymophisms could be supported in light of the "cost" if they were actually being selected for... But assuming an appreciable number of mutations need to undergo positive selection for, say, the emergence of a new mammal, it is still not altogether clear to me that the population sizes of mammals, particularly larger mammals, could shoulder the reproductive capacity required to sustain this selection. I suspect that recombination, epistatis, etc. go a long way towards making these scenarios more possible, but clearly there is still a lot we don't know, and it isn't clear when or if we will ever know them. For instance, what is the lower bound on the number of mutations it takes to generate the average new species? How many of these will be selective vs. neutral/nearly-neutral? What sort of population size is necessary to allow this number of selection events? What is a reasonable timeframe for this to happen in? On the theoretical side of things, there seem to be several questions left lingering that we--or rather someone with more expertise than myself--should give more attention to. If they have been adequately dealt with, please provide refs as I would find them very helpful. Otherwise, numerically-based arguments about what evolution simply *can't* accomplish go a long way towards recruiting educated folks--engineers, etc--to the other camp. Clearly a lot of the relevant parameters are difficult if not impossible to pin down. But that doesn't stop them from crafting arguments with "reasonable" parameters that inevitably show evolution is hopelessly unfit for the task at hand. To then respond, "we just can't know that well enough to make a proper estimation" surely comes off as evasive and "hiding behind ignorance" to much of the public. There are some matters, perhaps, about which we must be more proactive. For us here, the battle may be long over, but clearly on mainstreet and in the political arena it's very much alive and kicking. Any thoughts? Should we perhaps put more energy into patching up some of the loose ends that--while they may not seem all that interesting from the standpoint of research program--are nevertheless crucial in strengthening the credibility of evolution in the eyes of the general public??

By grrreat_ape (not verified) on 04 Jun 2006 #permalink

I'm going to try to address some of grrreat_ape's questions. I'm not too familiar with Haldane's dilema. David B provides a description of it at GNXP. I can't say how good this description is because of my poor understanding of Haldane's orignal argument. From what I can gather, it's become quite a popular topic amongst creationists.

On to some of the things I can address:

For instance, what is the lower bound on the number of mutations it takes to generate the average new species?

One. We can cause good species to produce viable hybrids using organisms with a single mutation.

How many of these will be selective vs. neutral/nearly-neutral?

The Dobzhansky-Muller model allows for allopatric speciation to proceed without selection. Coyne and Orr, on the other hand, present evidence that speciation (even in allopatry) requires divergent selection.

What sort of population size is necessary to allow this number of selection events?

Some folks argue that speciation occurs when small populations become isolated from a parental population. A founder event means the isolated population may not have alleles at the same frequency as those in the parental population. Also, because the founder population is small, mildly deleterious alleles may fix, allowing the population to overcome a fitness trough that was insurmountable in the parental population.

The short answer: speciation can proceed by many different routes, so there is not single answer.

What is a reasonable timeframe for this to happen in?

One million years? From what I can gather, a lot of people think it takes about a million years for speciation to proceed from start to finish. I don't know where this number comes from.

On the theoretical side of things, there seem to be several questions left lingering that we--or rather someone with more expertise than myself--should give more attention to.

Um, you never got around to saying what those questions are.

As for your later points, I don't think creationists can be spoken to rationally. Some may be intelligent people, but they have shut of part of their mind to reality to allow them to believe in their sky fairy. You can throw all the evidence you like at them, and it won't do a damn thing to change their opinion.

I was under the impression that Ewens (1972, amnat 106, 273-282) took care of Haldane's dilemma once and for all, but I might be wrong. The paper hasn't actually been cited that often.

RPM, Ido,

thanks for the helpful responses. I'm reading through some of Ewens work now, and it is definitely pertinent to many of these issues. I'll have more to say a bit later when there's more time for typing.

By grrreat_ape (not verified) on 05 Jun 2006 #permalink

Billy D (no, not that Billy Dee) is pimping an anti-evolution book by John Sanford.

Hmm.

Filed under: Intelligent Design -- scordova @ 1:51 am

Hmm.