Sunday, December 31, 2006

Human Genetics round-up   posted by agnostic @ 12/31/2006 12:19:00 PM

The new issue of Human Genetics has three articles that may interest GNXP readers. A study on the genetics of height looked at two groups of normal Japanese males (i.e., no cases of Marfan Syndrome), one tall (+2 SD; N =219) and one average (+/- 1 SD; N = 209). A SNP in the FBN1 gene was significantly overrepresented in the tall group, and the number of copies was significantly positively correlated with height within each group too. The mutation doesn't result in an amino acid change.

A study on the genetics of skin color variation looked at loci from Europeans, Africans, and Chinese (HapMap and Perlegen data) and performed tests of Fst (a measure of within-group vs. between-group variation) and decay of haplotype homozygosity (a measure of positive selection similar to the integrated Haplotype Score statistic used in the Voight et al 2006 study). They largely confirmed previous findings, but they also discovered a new locus involved in skin color differences between Chinese and non-Chinese populations, DCT (p. 617, all square brackets mine):

The core SNP was chosen as the SNP within the DCT gene with the most extreme riEHH value [i.e., their measure of positive selection]. This SNP, rs2031526, has a riEHH value within the top 1.7% of the riEHH distribution and has Fst values within the top 0.6 and 0.3% of the Afr–Chn and Eur–Chn HapMap Fst distributions respectively (AC Fst = 0.718; EC Fst = 0.607). Figure 2 demonstrates that the derived A allele in the Chinese is found on a high frequency haplotype with long-range homozygosity, while the Europeans and Africans show substantial breakdown of homozygosity over the same physical distance on the high frequency, ancestral G allele haplotype. Similar results were obtained when other core SNPs with extreme riEHH values from the DCT gene were used (data not shown). Thus, the DCT gene harbors a signature of local positive selection in Chinese using both Fst and LRH-based tests [LRH = Long-Range Haplotype], and is therefore a potential candidate to account for the differences in skin pigmentation between the Chinese and other human populations.

So here we have yet another example of different populations converging on more or less the same phenotype (lightish skin) via somewhat different evolutionary genetic paths. UPDATE: Dienekes links to another new study of convergent evolution of skin color in Europeans and East Asians.

Lastly, a study on the make-up of neutral portions of the African-American genome examined how much Europeans and Africans have contributed. The gist, including numbers, is contained in the abstract. In brief, they found sex-biased gene flow: European male - African female pairings are primarily responsible for admixture, a pattern which the authors note is also common in Native American - European admixed populations in the Americas. And apropos of a recent query at Razib's ScienceBlog, here's what they found on the Native American contribution to African-Americans:

A small contribution from Native American and Asian populations to the founding of African Americans has previously been reported (Parra et al. 2001; Smith et al. 2004; Reiner et al. 2005). In those analyses, the genetic contribution from Native Americans and Asians in individuals is at most 2.6% and generally falls between 1 and 2%. Reiner et al. (2005) showed that there was a greater likelihood that the African American population descended from two populations (as opposed to one, three, or four) which was consistent with our own STRUCTURE analysis, where two populations fit the data best. These analyses showed that it was not necessary to include Native Americans and Asians in the founder populations. Our Y chromosome analysis also supported that there was little contribution from Native Americans since the predominant Y haplogroups found in Native Americans P(xR) were not observed in this sample of African Americans.