Selection & African Americans

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

I already posted on the new paper on African American Genetics. I noticed that Frank Sweet says:

It is interesting that the 18 percent mean of Euro DNA markers in A-As has been holding steady for about 8 years now, having replaced the prior estimate of 25 percent.

Where did the prior estimate come from? I recall seeing it as well too. Were the older markers biased towards ones which might have been shaped by recent selection? The new paper doesn’t have anything definitive in regards to this (they they mention the variance in African vs. European across different regions of the genome), though certainly some genes which affect malaria seem to have been shifted away from what you’d expect.



  1. John Relethford has a short discussion of the admixture fraction in his book, Reflections of Our Past, pp. 214-218. Most of this can be accessed on Google Books (search in book for “Parra”). The admixture fraction varies a lot regionally in the U.S. and there is an urban/rural difference. Also a big paternal/maternal difference — so much that a switch from classical markers to whole-genome (including X) might have depressed the admixture fraction a bit. 

  2. There must have been estimates based on blood types. Just take European distribution, African distribution, and figure out what level of admixture would produce the African American distribution. This would work quite well, as long as there is no selection for blood type (which I think there was!). So, maybe that produced the 25% estimate?