Saturday, April 21, 2007

Alleles on the move?   posted by Razib @ 4/21/2007 01:03:00 AM

Hsien Hsien Lei reports on an interview in the May/June issue of Stanford Magazine with L.L. Cavalli-Sforza:
...A major genetic change which started already some centuries ago, with the navigation of the oceans, and is becoming faster now, is globalization. This is having major genetic consequences. It will bring back greater unity of the species, by diluting and eventually canceling differences among ethnic groups existing today, that are largely if not exclusively the consequence of adaptation to environments that differ most climatically to which modern humans spread in the last 50,000 years....

One of the novel insights of the field of human genomics is that universal selection pressures have resulted in alternative genetic responses to converge upon the same phenotype. For example, the light skin of East Asians & Europeans are independent derivations from the darker skinned ancestral type, and the genetic architecture of the trait attests to this. Now, the extent of intergroup admixture today is actually very modest, after all, very little European specific genetic material is entering the populations of China or South Korea! Nevertheless, in places like the United States widespread admixture between Europeans & East Asians is occurring, so you have the confluence (collision?) of genetic architectures which emerged in parallel meeting their "other half" for the first time. Since East Asians and Europeans have already been subject to sweeps which moved them (presumably) to their phenotypic optimum the selective value of the introduced alleles is probably not that high. On the other hand, pleiotropy implies that many loci have myriad side effects, and one never knows what can happen when you jump from one genetic background to another (good or bad). Since many of the alleles which differentiate East Asians from Europeans that evolved in parallel are the result of relatively recent sweeps I doubt they're embedded in essential contingent coadapted gene complexes, so I would be surprised by widespread negative or positive fitness effects because of statistical epistasis.

The reason I'm focusing on alleles with selective value is that moments like the settlement of the Americas by a small group of Iberian men and the generation of a mestizo population within a century through hybridization are rare events. If random-mating forces are what we depend on to eliminate between group variation I think we'll be waiting a long, long, time, because people tend combine in pairs of likes. It might only take one migrant per generation between demes to keep them from wandering off into alternative genetic directions (the larger the population the smaller the between generation sampling variance, while the smaller the population the bigger impact that one migrant can have on the gene pool), but many demes haven't had that much migration for a long, long, time. In contrast to a focus on neutral loci and total genome content, Loren Rieseberg's work has suggested that the spread of high advantageous alleles can maintain species continuity and coherency.

Related: 10 questiosn for L.L. Cavalli-Sforza.

Labels: ,