Saturday, February 02, 2008
I've been blogging the HERC2/OCA2 story a fair amount. It seems this genomic region is the locus of main effect for variation of eye color in Europeans, in particular blue vs. non-blue eyes. But I also pointed out that this locus has also been connected to variation in skin color, and while that variation is additive in effect, the variation on eye color exhibits strong dominance/recessive dynamics. My inference here is that it is more plausible that selection occurred on skin color, while eye color was a tissue specific expression pattern which emerged as a byproduct. Peter Frost has an objection to this:
The correlation between eye color and skin color may simply be an artefact of geographic origin. Europeans vary clinally for both eye color and skin color along a north-south and west-east gradient, so if the pool of subjects is geographically heterogeneous you will almost certainly get a correlation between eye and skin color. But this doesn't prove a cause and effect relationship.
Fair enough. Spurious associations driven by cryptic population substructure is one of the main reasons Structure was developed. I responded to Peter here, here and here. The short of is that I don't know of any analysis within an admixed population like African Africans, which would settle the matter, but there are plenty of other points which would suggest that we should look at the skin color trait (and, to be fair, if substructure exists at the level of British Isles origin samples we really need Strucure!).
But there was something that has been bothering me: eye color difference exhibits a lot of dominance/recessive dynamics in expression. The skin color data here does not, and aside from KITLG (which is dominant for light skin) all the other loci seem additive and independent (the report of epistatic effects here & there don't seem reproduced very often). One of the main reasons that I am favoring a skin color model as the phenotype driving selection is that if it is additive it is exposed to selection immediately at low frequencies. In contrast, recessive traits at low frequencies have the problem that most copies of the allele which increases fitness are still in heterozygotes which mask them from selection. It came to my mind that the different assumptions about dominance would matter in terms of long term evolutionary dynamics and how that would be realized in terms of results from tests for selection. So I found this paper, Directional Positive Selection on an Allele of Arbitrary Dominance. It says:
...fixation of a beneficial allele leaves a signature in patterns of genetic variation at linked neutral sites. If this signature is well characterized, it can be used to identify recent adaptations from polymorphism data. To date, most models developed to characterize the effects of positive directional selection (termed "selective sweep") have assumed that the favored allele is codominant. In other words, if the fitnesses of the three genotypes are given by 1, 1 + sh, and 1 + s (where s is the selection coefficient), then h = 1/2....
For skin color h would be 1/2 for HERC2/OCA2, it has half the effect on the trait value. Assuming proportional selection based on the character value two copies would be better than one copy which would be better than no copies. In contrast, for eye color the h would be between 0 and 1/2, and probably closer to 0 because of predominant recessivity in expression for blue eyes. That means the fitness of those with one blue eye copy would be much closer to those with no blue eye copies than those with two; to the homozygote recessives would go all the benefit. On to the results:
...when h is small, most of the sojourn time is when the allele is at low frequency in the population. During this phase, the allele will have the opportunity to recombine onto other backgrounds. In other words, the favored allele will tend to increase in frequency on multiple backgrounds, preserving more of the diversity that existed when it first arose. In contrast, for dominant alleles, most of the sojourn time is spent at higher frequency, when there is less opportunity for the favored allele to recombine onto other backgrounds. This results in a wider signature of a fixation event for larger h-values.
Why the bolded parts? From A Map of Recent Positive Selection in the Human Genome:
Some of the strongest signals of recent selection appear in various types of genes related to morphology. For example, four genes involved in skin pigmentation show clear evidence of selection in Europeans (OCA2, MYO5A, DTNBP1, TYRP1). All four genes are associated with Mendelian disorders that cause lighter pigmentation or albinism, and all are in different genomic locations, indicating the action of separate selective events. One of these genes, OCA2, is associated with the third longest haplotype on a high frequency SNP anywhere in the genome for Europeans....
I don't know if my connection of inferences here is valid, and the paper I originally referenced makes clear that it is important to frame these sorts of assumptions within their statistical context; just because something is less likely does not mean it is impossible. I've sent out emails about OCA2 and skin color, and will report back, but at this point I suspect that the final proof in the pudding will have to be admixture analysis in a group like African Americans. But I think the above makes it more likely that whatever was going on 10,000 years ago did not express as a recessive phenotype.