I'm trying to understand your 59 alleles argument against a recent
bottleneck in the human population. Please help me out. Two points are
unclear to me. i) the meaning of the word generation, and ii) how common
each of the 59 alleles are.
You wrote:
>There could only be 10 alleles in the genome (forget a.a.) of the people who
>got off the ark. If they were the entire human population at that time, then
>there were only 10 alleles in existence for the MHC complex anywhere in the
>universe. So the question is, "How long does it take for the other 49
>alleles to arise by mutation from just these 10?"
Are these alleles all embedded (I mean common) in the human population? Or
if we were to sample more people, might we find many more? The importance of
this question will become apparent below.
>In a previous post I gave the rates of mutation of 1 x 10^-7 for each
>nucleotide location each generation.
This rate seems impossibly high if by each "generation" you mean all the
kids of all the people (which seems to be what you mean by generation
below). If it is the rate for parent to child, then fine. But just because
there is a mutation from parent to child doesn't mean that it will become
common in the population. If the population is N when the mutation occurred,
then isn't the odds of finding that mutation 1/N for *all* succeeding
generations? So, I would presume that if there are 59 common alleles, and
they arose by mutations from an original 10, then they must have arisen very
early when the population was very small, or at least in small subpockets of
civilization that eventually became prominent. (I'm wondering if this kind
of reasoning is what makes Terry Gray say the original human population was
on the order or 500 individuals?)
>If you have a 1000 unit long gene, then each generation has a
>10^-3 chance of having one mutation in that gene. Thus in about 1000
>generations you would expect to have 1 mutation in that gene on average.
I'll frame the same question in a different way. Let's take the world's
population to be 5 billion. Let's say an allele has to be in 1/10,000 people
for it to make it into the class of 59. Then 500,000 people have that
allele. Are you asking me to accept that there is a 1/1,000 chance that
500,000 people will be born with the same mutation in the same generation?
If on the other hand, by generation you mean 1 child per parent, and the
commonality of the allele is not an issue, then the number of alleles in the
population clearly depends on its size, which somehow doesn't figure into
your calculation.
Jim
Jim Blake
Associate Professor
Department of Electrical Engineering
Texas A&M University
College Station, TX 77843