Friday, July 15, 2005

Correction on MC1R....   posted by Razib @ 7/15/2005 08:08:00 PM

I misreported Heather Norton's presentation in the post below. I emailed her a few questions, and she responded:

Actually, my data do not show that Europeans and East Asians have different MC1R alleles, but rather that they appear to be different for two alleles in the genes TYR and MATP. I didn't see strong differences between Europeans and East Asians at the two MC1R SNPs that I stated that I had presented them at a conference in Europe. The conference was in fact at the Annual Association of Physical Anthropologists Meeting in Milwaukee, Wisconsin earlier this year.

...I'm in the process of writing up the results of the work that I presented at the AAPA meeting for publication I am not prepared to discuss them fully right now, but I can tell you that I think that MC1R is often attribted a larger role in pigmentation regulation than perhaps it really has. In no way can I disupte that MC1R sequence variation is highly constrained, especially populations living in regions of high UVR (like West Africans), and I do believe that this is because certain MC1R mutations lead to lighter skin, which would be maladaptive in these regions. However, pigmentation is a complex trait that is under the control of many genes, and I think that while MC1R might explain a lot of the variation in skin and hair pigmentation phenotypes that we see among different European populations it is not so great at explaining variation at more global level. In other words, I believe that there are a number of other genes that may play just as important a role, if not a greater role, than MC1R in explaining these differences (and of course, I'm saying this because of my some of my results do indicate that other genes are involved). I think that one reason why MC1R is often cited in discussions of pigmentation differences is because it is a short gene (only 1 exon ~ 1kb in length) and so is very easy to sequence. Also, some MC1R alleles have very easily visible effects on phenotype (these would be the strong RHC alleles), which of course makes things easier. So, I'm not knocking MC1R, I think it's a very important gene and I think that it's great that it's been so well-studied. However, I don't think it tells the whole story about global
human pigmentation variation....

She can't answer my direct questions right now because there is a paper pending publication. Here is her abstract from the presentation:

Using measures of locus-specific differentiation to find genes underlying traits subject to recent genetic adaptation: a test case using skin pigmentation.

H. Norton1, R. Kittles2, C. Bonilla2, J. Akey3, M. Shriver1. 1Department of Anthropology, Penn State University, 2College of Medicine and Public Health, Ohio State University, 3Division of Human Biology, Fred Hutchinson Cancer Research Center, Seattle, WA.

A number of DNA sequence-based statistics are available to identify signatures of natural selection. However, sequencing large numbers of individuals across multiple genes can be costly and time consuming. An alternate method that uses allele frequency data has received less attention, but may be more efficient for large screening studies. This method is based on the idea that demographic events affect loci across the genome equally, while adaptation affects individual genes and nearby markers. We have applied the locus-specific pairwise FST (lspFST) to survey seven pigmentation candidate genes from six geographically diverse populations. Using the allele frequencies at these genes, we calculated the lspFST statistic and compared it to an empirical distribution based on 11,078 SNPs analyzed in the same populations. With this comparison we are able to take into consideration the demographic histories of the populations and calculate likelihoods of the data given neutral evolution. Several pigmentation candidate genes show evidence of non-neutral patterns of differentiation. Interestingly, population differentiation at pigmentation
candidate SNPs was observed both for populations differing in pigmentation phenotype (ASIP and OCA2), as well as for populations similar in pigmentation phenotype (TYR). Additionally, SNPs in MATP show high levels of European-specific population differentiation. These results suggest a strong role for natural (and/or sexual) selection in shaping human pigmentation variation. Patterns of allele frequency and lspFST variation at TYR and MATP between Europeans and East Asians raise the possibility that natural selection may have acted on different alleles to produce a similar adaptive phenotype in these populations.