what does a high derived allele frequency mean?

Question

[epigene]

I read in a paper something on allele frequency that says:

We identified differences between human genomes and the inferred human-chimpanzee ancestral genome where humans carry a derived allele with a frequency of at least 95% (14.9 million SNVs and 1.7 million indels). Nearly all of these events are fully fixed in the human lineage, with fewer than 5% appearing as nearly fixed poly- morphisms in the 1000 Genomes Project variant catalog (derived allele frequency (DAF) ≥ 95%).

I don''t have much background on allele frequency and evolution theory. I'm curious what does a high DAF mean? Does it mean that between human and chimps, most derived alleles are now fixed (as they are over 95% meaning 95% of the human population have a particular derived allele)? And DAF of 95% doesn't suggest anything on whether there is positive or negative selections on derived alleles? Is my understanding correct?

Thank you!

Answer 1

[Gabriel R.]

I'm curious what does a high DAF mean?

High derived allele frequency means that a mutation likely occurred somewhere on the human lineage and is now found in about 95% of humans. The underlying mechanism is unknown.

Does it mean that between human and chimps, most derived alleles are now fixed (as they are over 95% meaning 95% of the human population have a particular derived allele)?

No it does mean that just because some humans have a derived variant that all of them are fixed.

And DAF of 95% doesn't suggest anything on whether there is positive or negative selections on derived alleles?

No, a DAF of 95% could be due to:

1. chance, alleles will fix or be weeded out by chance especially if your effective population size is low. Chimps have a greater effective population size but could still fix or weed out variants by chance.
2. positive selection
3. weak background selection
4. error in the inferred allele for the chimp/human. It also implies that there was a single allele back then. It's possible that the one fixed in humans and the other in chimps by pure chance.

[epigene]

Thanks for your answer!

I'm a bit confused by "No it does mean that just because some humans as a derived variant that all of them are fixed."

Can you clarify on it?

Why do chimps have a greater effective population size than human?

[Gabriel R.]

Sorry that should read "some humans have a derived" I didn't write this, the pre-caffeine version of me did :-)

I mean by this that if you find a variant in the 1000G files and assume an infinite sites model (no back mutations or tri-allelic sites), either the reference or the alternative is ancestral and the other one is derived. Most mutations in there are only shared by a small number of individuals. It's not because a variant appeared in humans that it is automatically fixed.

For the chimps, it's a broad statement to say that all chimps have a higher effective pop. size, read this article for discussion about the effective pop. size between great apes:

http://www.nature.com/nature/journal/v499/n7459/full/nature12228.html

[epigene]

So a human allele with high DAF does not always mean it's fixed in human?

Thanks for the paper link.

[Gabriel R.]

High DAF just means high frequency. Depends how you define "fixed" versus "almost fixed".