# How much of the genetic differences between humans and chimpanzees are due to genetic drift?

Both natural selection and genetic drift play a part in changing our DNA over time. Are there trustworthy estimates how much of the differences over the whole genome between us and chimpanzees are due to genetic drift and how much is due to natural selection?

• Interestingly, the rate of accumulation of new neutral mutations is independent on the population size and is equal to $\mu$ (per lineage), where $\mu$ is (also) the mutation rate at a given locus. Mar 9, 2016 at 16:41
• @Remi.b : Let's focus on the whole genome. Mar 9, 2016 at 16:45
• Note that most selection pressures will actually prevent divergence between two sister lineages as they act in the same direction. Only cases where selection differs between lineages will cause divergence. This type of selection is simply called divergent selection. Mar 9, 2016 at 16:48
• @Rodrigo The loss of polymorphism through time (which is how one would quantify the concept of genetic drift) is higher in smaller population. Have a look to why-is-the-strength-of-genetic-drift-inversely-proportional-to-the-population-size. Mar 10, 2016 at 0:50
• @Rodrigo However, the rate at which neutral mutation fix is the same. The reason is that the probability of a neutral mutation to fix is $\frac{1}{2N}$. There are $2N\mu$ new mutations entering the population each generation and therefore the rate of fixation of neutral mutations is $\frac{1}{2N} \cdot 2N\mu = \mu$. This classic result is essential for phylogenetic reconstruction typically as we can infer time of divergence from the rate of neutral divergence without knowing anything about $N$. Mar 10, 2016 at 0:50

The easiest way to track whether a particular protein-coding gene has diverged between two species because of drift or actively maintained selection is to look at the ratio of synonymous to nonsynonymous mutations within that gene. A synonymous mutation is one that doesn't change the identity of the amino acid within that protein; a nonsynonymous mutation is one that does. Synonymous mutations are neutral by definition; their presence gives us an idea of how much time has diverged and how frequently mutations in that protein coding region change over time.

A dN/dS (Nonsynonymous mutation to Synonymous mutation) ratio of 1 suggests that the primary driver of change between species is purely neutral change. That usually means that it doesn't matter what is going on with that protein and is a little unusual--I would primarily expect to see something like that in a gene that is on its way to becoming a pseudogene.

A ratio of less than one means that synonymous mutations are much more likely to persist in that region than nonsynonymous ones. This is usually a sign that purifying selection is happening--for whatever reason, it's important to the function of both species that those proteins stay relatively similar.

A ratio of more than 1 means that there is diverging or directional selection happening, and that that protein type is changing between the two species because selection is actively modifying one or both proteins.

Here is a paper specifically analyzing whole human and chimpanzee genomes using this metric.

• You could improve this answer if you summarize what the linked paper concludes (in general, answers on StackExchange should stand on their own; links and references provide support, but we hope for answers here to outlast links that could change or go dead in the future). Jul 31, 2017 at 21:16
• "Synonymous mutations are neutral by definition" that's wrong. It doesn't change the shape of the protein but it does change protein synthetisation speed/amounts. Aug 2, 2017 at 7:17
• @Christian It's wrong in so far as it is a simplification/assumption, and a useful one at that. Sep 29, 2017 at 21:54
• @canadianer : I see no reason why there shouldn't be evolutionary pressure to increase or decrease the synthesis of a certain protein and that's what a synonymous mutations does. Sep 29, 2017 at 22:20

Wildman et al. (2003) investigated 97 genes in humans and chimpanzees and show evidence that 30 of them (~31%) have evolved under positive selection. This is not per say an estimate of what fraction of substitutions are adaptive but it gives an idea.

Such estimates can be quite hard to get. I don't know any such estimates between human and primates but I know one on Drosophila.

Eyre-Walker and Keightley (2009) that about 40% of all substitutions between Drosophila melanogaster and D. simulans are adaptive.

Population sizes being larger in the Drosophila lineage than in the Hominidae lineage, I would predict that the fraction of substitutions that are adaptative would be significantly lower between humans and chimpanzees (prediction which agrees with Wildman et al. (2003)).

So, to the question How much of the genetic differences between humans and chimpanzees are due to genetic drift?, I would answer: More than 60%! But remain cautious realizing this is a vague estimate.