From the conclusion:

However, a caveat should be considered when interpreting the differences between African-American and non-African populations: the statistical power to detect selective sweeps is likely to be much lower in the African-American sample. Because the CLR test is based on a complete sweep model, the recent admixture of African and European lineages in the African-American population probably weakens the signal of Africa-specific selective sweeps.

As we used to say when I was a kid, "no duh."

So why do investigators use African-Americans, rather than Africans, in these studies? Is African DNA really that hard to get? Or is there some... gasp... political reason?

So why do investigators use African-Americans, rather than Africans, in these studies? Is African DNA really that hard to get? Or is there some... gasp... political reason?

there's not a conspiracy everywhere, man. the HapMap did indeed use an african population. but in their effort to type every single SNP they could get their hands on, they didn't use a constant ascertainment scheme, which makes their data harder to use for population genetic studies (especially those, like this one, that use the site frequency spectrum in their calculations).

this paper is based off a different data set, one which used a uniform ascertianment scheme. It was collected by Perlegen, an American company that decided it was probably easier to get African-American samples (and also, since they're going to look for disease susceptibility alleles in the USA, it might be a more useful sample for them).

A Fearless Champion of Truth said that African Americans had a 25% admixture of white genes in Race Differences in Intelligence.

[this blog is not a forum to rehash a 3 year old dispute with abiola lapite]

Edited By Siteowner

African Americans had a 25% admixture of white genes

European admixture in African-Americans varies a lot by individual and by region in the US. see here.

25% is a little on the high end.

Look, I respect Lynn's scholarship, so could I still call Lynn that although Abiola was most certainly sarcastic. I will say that Lynn's work is the apogee of scholarship!

I'll quote from page 70 of Race Differences in Intelligence: "in Southeastern states the percentage of European genes among African-Americans is quite low. For instance in South Carolina it has been estimated at 6 percent (Workman, 1968) and in Georgia at 11 percent. These admixtures of European genes should increase their IQ [from the genotypic IQ of 80 from an African black with no white admixture] by 1.2 to 2.2 IQ points, respectively, giving them an IQ of 81.2 and 82.2. This prediction has been confirmed by the study of 1800 African Americans in five Southeastern states by Kennedy, Van der Rei, and White (1963), which found their IQ on the 1960 Stanford-Binet was 80.7."

[early comments should stay on topic. last warning]

Edited By Siteowner

p-ter,

Thanks - seems perfectly sensible.

I can't help thinking, though, that the authors aren't too unhappy about having to include that caveat in their conclusion. Which could otherwise be rephrased in rather distracting terms. Not that they are part of "The Plan," but I think most sensible scientists, especially in this area, are political in that they want to stay as far from politics as possible.

mencius,

you're right that scientists in this area often want to avoid politics. but the caveat in the conclusion of this paper is absolutely necessary from a scientific standpoint. some papers have reported less recent selection overall in African populations, while others have disputed that (see the first paragraph of the discussion in the Voight et al. paper I linked in the post for a little commentary, it's open access). the authors of this paper need to include their caveat because their data honestly cannot be taken as evidence one way or another. it's included for scientists, not for the general public or for political reasons, as far I as I can tell.

Well, it seems that it is possible obtain an African American sample with minimum admixture if you know where to look. I am sorry if the IQ derail is off topic, but it seems that African Americans in the Southeastern states have lower white admixture. Although IQ might be off topic, I thought it was relevant to add that the estimatation of white admixture in certain African American populations is commensurate with the results of psychometric testing.

Regarding hemochromatosis, it has low penetrance (see: http://scienceblogs.com/gnxp/upl...enicspaper.pdf) so it wouldn't be selected out easily. But does anyone have the frequences for the related genes.

Well, it seems that it is possible obtain an African American sample with minimum admixture if you know where to look.

obviously. but even NIH grants go only so far....

re: admixture. the data in this paper was generated by perlegen ultimately for use in mapping disease alleles. yes, it was used in this paper for popgen analyses, but that's not its raison d'etre. having lower admixture reduces power in the popgen analyses, but for admixture mapping of disease alleles, there's a certain amount that's desirable-- less is definitely not uniformly better.

something to keep in mind-- the huge datasets people use now in popgen were generated for medical reasons, then twisted around to be made usable in popgen. something that makes population genetics more difficult might make medical applications easier. so not everything is going to be ideal.

Human evolution is continuing, and making us genetically different.

Evolution is on, but human speciation is? Nowadays, homogeneization may be a stronger force.

I have no data this or other way either.

Nowadays, homogeneization may be a stronger force.

well, i don't know 'stronger' or 'weaker,' but probably enough between-deme migration to prevent speciation.

something to keep in mind-- the huge datasets people use now in popgen were generated for medical reasons, then twisted around to be made usable in popgen. something that makes population genetics more difficult might make medical applications easier. so not everything is going to be ideal.

...you sure that wasn't done for political reasons???

i read a review and several abstracts on heat shock proteins and looked up what the different selected ones were. i came to the conclusion that if i could come up with an explanation for the Hsp changes in an evening of casual pubmed flipping then the authors probly woulda found it in the months preceding publication.

you can broadly say they are involved in processing proteins, but i think you'd be hardpressed to find a cellular process that some Hsp doesn't have its fingers in. maybe you can press upon Daily Transcript to enlighten us..

p-ter,

I absolutely agree, and if I didn't your argument would convince me :-)

I'm just saying that I suspect the first thought on the authors' minds might not be "now let's find us some pure African DNA, so we can put that article in the Charleston Times-Overseer 'bout how the Nigra people is really less evolved than us white folk."

Not of course that a lower rate of selective sweeps means anything of the kind. Just that, in a world where there actually was a Charleston Times-Overseer, it would be taken that way. There isn't, but people are still a little sensitive on the subject. I don't think it's a negative reflection at all on these researchers to suggest that they might not be 100% eager to go there.

Nah, the political implications will manifest themselves if such as genomic scan detects a selective sweep for an allele that is correlated with g that happens in some population groups but does not occur in others.