Direct estimation of mutations in great apes reconciles phylogenetic dating

A sudden change in mutation rates would need to be explained. It definitely doesn’t seem to have been expected or predicted in the case of human evolution. Quite the contrary.

If change in rate is likely as you say. Then there is no basis to calculating divergence times over millions of years assuming the same mutation rate.

There is also purely human data suggesting that mutation rates in our lineage were faster in the past, such as this study:


Of course there’s basis for it, as long as researchers are aware of the limitations of using a single mutation rate. They are, which is why such divergence date estimates are often associated which pretty wide confidence ranges. The more recent the divergence the more precise we can usually be.


I am making the same point. You can’t make both claims at the same time-
a) Mutation rates are stable enough to give good approximations of divergence time.
b) Change is mutations is likely and hence the need to change mutation rates for human beings to correct the divergence time calculation is within expected norms.

If “b” is true, then “a” is false and vice versa. I am perfectly willing to see “a” as true and the human case to be an exception. This seems to be the point you are making.

A value of 1.61 per generation would be equal to 0.64 mutations per billion years… this is similar to the mutation rate of chimps in the paper.

That’s definitely interesting.

Do you think it is inconsistent for me to simultaneously believe that mutation rates are generally stable enough to give good approximations of divergence times and that accounting for changes improves the estimates?


No, it gives a rate of 0.55 per billion years, given a 29-year generation time (the OP paper mentions this). Intermediate between modern human rates and modern great-ape rates (inferred to also be ancestral rates).

1 Like

Didn’t think about it.

Depends on whether you can account for changes directly. Otherwise it’s not very different from calibrating tot he fossil record.

Ok… I used 25 years.

So that makes it a difference of about 2.7 my…

Yes, did you expect anything different? Varying critical parameters will affect inferences, that’s why so much effort is put into gathering data about those parameters to make the best estimates. It’s also why it’s common to use fossil evidence to calibrate trees, as they offer useful anchor points when trying to estimate divergence times. Not to say that fossil calibration is flawless or without its own uncertainties, but it helps.

1 Like

It’s exactly what I expected . Afterall it’s not a complicated equation.
I actually appreciate the direct measurement approaches as they give concrete data.

It seems clear that earlier expectations of mutation rates calculated based on the species divergence between humans and chimps from the fossil record were wrong.
So it’s definitely an interesting paper. And if the lower mutation rate gets established, then we will have the interesting result that assumptions of a nearly constant mutation rates doesn’t hold in all cases.

It will be interesting to see if this will become the general trend as more genomes get sequenced.

We already know that this assumption doesn’t hold in all cases. As I already said, that’s one reason why there are such large confidence ranges associated with molecular clock dating, especially for deeper nodes that are ancestral to a greater number of lineages. For example, here’s a time-calibrated tree of musteloid evolution from a 2018 paper:

Look at those wide 95% confidence bars (blue).


Nope, because of deep coalescence. The values are with in the range of the absolute values we expect. The relative values of different types of mutations, as @glipsnort has explained, are stupid close to the correct values.

1 Like

I didn’t understand. How is a difference in predicted divergence times connected to deep coalescence? Based on measured mutation rates of human and chimps, the divergence from CA for chimps is 9.46 Mya and for human beings is 15.7 Mya.

Can you elaborate.

I am not sure of this.
Take the paper you presented. Atleast one reason for the width in the timelines being long should be because of the incompleteness of the fossil record.

Even,If actual mutation rates are calculated for extant species directly and they show a large difference from earlier estimations, I don’t think this will change the time a species is supposed to have emerged. They will correct using a change in mutation rate because of parsimony… and the fossil record being a better indication of the time factor here (assuming the times of fossils are established using more direct/established methods).

Yes, that’s one reason, another is the one I mentioned. They use relaxed mutation clocks in these analyses, allowing for a range of mutation rates and changes along branches.

That depends on the phylogenetic distribution of the mutation rates, doesn’t it? If actual mutation rates are calculated for good set of animals in a clade, then the updated calculations may well change the timelines somewhat.

1 Like

Depends on how close the actual mutation rates come to the ones obtained based on the fossil record.
If variations of upto 50% are common it will cause a lot of hacks… and humans are a relatively recent species.