October 28, 2010

Very low mutation rate, mwahahaha!

I've been reading around about the publication of a paper on the 1000 Genomes Project and what seems most interesting is that the observed mutation rate in humans would be of 1.1 x 10⁻⁸, much less than usually accepted. 

John Hawks seems particularly affected by the repeated observation of this rate, which he had previously discussed in March (on another different paper), up to the point that he seems about to accept the reality of such a slow molecular clock. 


So...


Other elements of interest may be the finding of many previously unknown alleles, which, unlike those known earlier are rarer and found often only subpopulations, rather than wildly scattered through the planet.

Also that brain-related gene duplications may be at the root of human intelligence, when compared with other great apes (see Science News).

5 comments:

  1. I'm very interested in that mutation rate. Can you tell us more about what the consensus mutation rate in autosomal SNPs is estimated/presumed to be by geneticists? So we can see by how much of a factor they would be off.

    Now let's see if I got things right. This would mean that many studies that attempted to calculate an age for the origin of things like lactase persistence, or white skin in Europeans, would have to increase their dates by X multiple. And what about the regions of the autosome that are already believed to have such an old coalescence date that some geneticists are wondering wether it means a small part of the genome was the result of mixing with non-homo sapiens?

    ReplyDelete
  2. This is not haploid but overall: that mutation rate was observed in two different family trios (an Euro-American and an Afro-American family) and is totally consistent with an older finding mentioned by Hawks (the paper is pay per view, sadly).

    I cannot specify, specially for haploid Y-DNA (the family trios were both with a daughter) but it's clear that the molecular clock estimates in general are falling very short of the real thing.

    "This would mean that many studies that attempted to calculate an age for the origin of things like lactase persistence, or white skin in Europeans, would have to increase their dates by X multiple".

    Yes, approximately double, but as there's no standard method, the variation should depend. Cautious estimates like Neanderthal-Sapiens divergence of 700 Ka, that already factored some 2 extra Ma in the Homo-Pan divergence, only need some 50% more, more typical ones may need to double or slightly more, while those Dienekes-style (pedigree rate) are so far from the mark that is not even worth correcting.

    My opinion anyhow.

    "And what about the regions of the autosome that are already believed to have such an old coalescence date that some geneticists are wondering wether it means a small part of the genome was the result of mixing with non-homo sapiens?"

    No idea. I don't know anything about that.

    ReplyDelete
  3. "And what about the regions of the autosome that are already believed to have such an old coalescence date that some geneticists are wondering wether it means a small part of the genome was the result of mixing with non-homo sapiens?"

    I'm sure that's exactly what it means.

    ReplyDelete
  4. The study makes a good case for that rate as a baseline in relatively ordinary circumstances.

    The notion that the mutation rate is truly constant and isn't influenced by environmental factors is considerably less well established IMHO.

    Indeed, factors as simple as advanced paternal age seem to have a great impact on the mutation rate in humans, and surely that must not be the only example.

    ReplyDelete
  5. @Andrew: Totally agree.

    I guess that if we could be certain that the observed mutation rate is consistent, we'd have a molecular clock of sorts. However if the observed mutation rate is as low as observed here and the other paper, current MC estimates are all wrong: they are all too recent (some maybe more than others).

    But still that cannot exclude periods and places when the mutation rate changed, for example as result of more elderly parents, local contaminants, accelerated evolutionary processes...

    And then another issue is how to account for demographic influences in how actual mutations survive and consolidate or do not. That's why the pedigree method is not generally accepted at face value... but corrections are tricky.

    For instance, I have been considering that large haplogroups (specially in mtDNA, where all or nearly all mutations are known for each line) may tend to "evolve" less because drift favors the large dominant clades (and that means that novel mutations have less chances of consolidating as new lineages. That would explain why very large star-like clades like M and H (which obviously experienced large and rapid demographic expansions at some point) also show less mutations downstream compared with "sister" clades like N or U respectively.

    ReplyDelete

Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (sorry, too many trolls).