Take the 2-minute tour ×
Biology Stack Exchange is a question and answer site for biology researchers, academics, and students. It's 100% free, no registration required.

Recently I have found myself in a discussion with a guy that isn't a creationist, but denies evolution. He cites several studies that show problems like this one, but I do not have the knowledge to refute it properly. Whats the problem with the following:

One problem is that population genetics shows us evolution destroys functional sequences much faster than it creates them in any large-genome animal with low reproductive rates, because deleterious mutations arrive faster than selection can remove them. This would include just about all mammals, reptiles, and birds--any of our would-be ancestors over the last 300m years.

Specifically, humans get 60 to 160 mutations per generation, 10-20+% of our genome is sensitive to substitution, and this gives us 6-32 deleterious mutations per offspring (most slight). Even with the low estimate of 6, every child is less fit than their parents and natural (or even artificial) selection can only choose the least degenerate each generation. For every generation where a mutation can be selected, random mutations destroy 5-32+ previously selected. Beneficial mutations like lactose tolerance (a jammed switch) take 1000s of years to appear and fixate.

If you know the genome size, mutation rate, and reproductive rate, you can calculate at what rate deleterious mutations arrive faster than they can be removed:

  1. "It has been estimated that there are as many as 100 new mutations in the genome of each individual human. If even a small fraction of these mutations are deleterious and removed by selection, it is difficult to explain how human populations could have survived. If the effects of mutations act in a multiplicative manner, the proportion of individuals that become selectively eliminated from the population (proportion of `genetic deaths') is 1-e^-U, where U is the deleterious mutation rate per diploid, so a high rate of deleterious mutation (U>>1) is paradoxical in a species with a low reproductive rate." High genomic deleterious mutation rates in hominids, (Nature, 1999)

They worry about a deleterious rate of U much greater than 1 as being prohibitively high and we're talking about it being 6-32+, since we now know the functional genome is much larger than we thought it was in 1999. Taking their Poisson probability distribution and using U=6, that means 1-e^-6 = 99.752% of the population will have to be selected away each generation for one lucky enough to have no deleterious mutations. For two to survive and maintain constant population size that would require on average 806 offspring per female--impossible for most mammals, birds, and reptiles. Actually much more since natural selection is not omnisciently efficient.

share|improve this question

migrated from skeptics.stackexchange.com Jul 8 '13 at 14:06

This question came from our site for scientific skepticism.

This belong on Biology.SE –  Ilya Melamed Jul 8 '13 at 13:59
Biology.SE is actually very unhappy about creationism questions. However, I think this one’ll pass. –  Konrad Rudolph Jul 8 '13 at 14:05
Well, one problem with the calculations, disregarding the difficulty of defining "deleterious" quantitatively, is that we are diploids. Consequently each individual will be, to a first approximation, heterozygous at all of the new mutated sites. This will mean that most of them will have no discernible effect. The annals of human genetics are full of cases of asymptomatic 'carriers'. So it will only be in a rare homozygote that we get a potential effect that is strong enough for strong selection. –  Alan Boyd Jul 8 '13 at 15:41
@AlanBoyd could you expand on that into an answer? Maybe also talk about downstream effects - loss of heterozygosity, the fact that many genetic diseases still leave the bearer able to reproduce, etc. –  MattDMo Jul 8 '13 at 17:26
At a first brush, that seems either improbable or inconsequential. If the increased rate of mutation makes large genomes maladaptive overall, then natural selection could not eventuate them, but either A) natural selection eventuated them or B) chance did. B) is improbable; whereas A) is logically incompatible with the claim that mutation would make large genomes maladaptive overall. If they are not maladaptive overall, but merely have drawbacks (more mutations) then the costs outweigh the benefits, so that supports an evolutionary account of the origin of large genomes. –  Hal Aug 5 at 5:04

2 Answers 2

As @ AlanBoyd pointed out one of the problems that comes up in more complicated species (like plants, let alone humans) is that we are diploids. This serves to slow the changes that occur in the types of mutations described in your links (I will note the Nature article discuses several types of mutations not included in the referencing site).

I think the critical place to start here is that the science there, from 1999, is quite out of date. Not only have we now completed the genome, but we can make comparisons directly between human generations instead of estimations from chimps. This is partly because we've gotten much better at sequencing. I think a much better article to look at is from Keightley

That puts the mutation rate closer to 70 mutations per generation and a U of 2.2 using more classical methods. But that's probably not a good measurement ether. Quite simply as organisms get more and more complex, they often become more and more resistant to change. Being diploid is just a good example. Regulation from more complex proteins and exons is another. We don't mutate like a virus or bacteria because we have a lot of things going right and if we mutated randomly like that a lot of things would go wrong and quickly.

Instead we are finding that transposons are the rule not the exception in how we are adding variation.

I'm not sure how much detail you want me to go into, but here is a nice way to think about it. Large genomic organisms have endured a lot of fitness to gather the genomes they have, and they become more and more resistant to change after developing what works. This is of course a generalization and simplification, but is the general agreed pattern in evolution.

share|improve this answer
Also, because of diploidy, crossover during meiosis will get rid of a fair few of the deleterious mutations. –  Chinmay Kanchi Jul 9 '13 at 4:57
Absolutely. There are a long laundry list of mechanisms that have developed to slow/prevent deleterious mutations. Diploidy was certainly a major advancement (also in adding variation through fun things like sex). Do people think composing such a list would be useful? I could take the time to try and rattle of what immediately comes to mind, but I'm not sure such a list would be a good SE answer. I am still going through the bio-meta as I just discovered the forum a few hours ago. –  Atl LED Jul 9 '13 at 5:19

Also, not all parts of the genome respond in the same way to mutation. Non-functional stretches of genes will tolerate mutation readily (as an SNP may have no discernible affect on the gene's function - and will thus be 'silent'). Other areas of the same gene, such as DNA binding domains, might be crucial to the gene's ability to perform a certain function, therefore any mutation arising in that region would have a higher chance of producing a deleterious effect, and thus being eliminated by natural selection.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.