Wednesday, October 26, 2005

Africans & Europeans genetically distinct...

...if you go out enough decimal places and look only at Nigerians.

"across the 1 million sites examined, only 11 always contained a different genetic base in the Nigerian samples when compared with the same site in the European samples."

(I say "only Nigerians" because genetic diversity of sub-Saharan Africans is about as great as that which exists across all other peoples combined, last time I checked, consistent with the belief that modern humans originated in Africa and groups migrated out sporadically to populate the rest of the world)

P.S. A further elaboration as per request of my mom, who didn't understand the quote and doesn't like hyperlinks: Population geneticists have identified a few million SNPs (single nucleotide polymorphisms), which are positions in the 3 billion base-pair sequence of the human genome that are likely to be different if you pick two people at random. In recent news, researchers focused on 1 million SNPs, comparing them between individuals of different races and regions. Now what does it mean to say that two groups of people who look different or who live different places are "genetically distinct"? Minimally, it means (to me) that you would have better than even odds of guessing which group a person belongs to by looking at his or her SNPs. In an imaginary world of two equally populous groups, say orange and blue, you'd have this kind of genetic distinctness if 90% of oranges had a G at position 101 and 10% had a T, while blues were the opposite (90%T, 10%G at SNP 101). In this imaginary world, when DNA from the crime scene shows a G at SNP 101, odds are your murderer is orange. But suppose that in this world oranges and blues are really darn distinct genetically. Then that SNP 101 actually is a rather rare one, in that it forces detectives to guess or make a bet. If oranges and blues are really darn distinct though, the identity of nearly any SNPs enables them to say with certainty what color the murderer was. They'd still have to look specifically to the SNP positions, because the genomes of both groups are 99.9% identical. But at the SNP positions--or at most SNP positions--the identity of the base would be telltale.

Not so in a world of Nigerians and Europeans. Only 11 out of a million SNPs are telltale (0.001% of SNPs, 0.000001% of the genome overall). So I wouldn't call Nigerians and Europeans really darn distinct gentically.

No comments: