I was wrong about skin color in Europeans and East Asians

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

In December I posted a hypothesis about skin color evolution in East Asians and Europeans in light of new research which suggested that Europeans were fixed for a particular locus which explained a great deal of the variation between themselves and Africans. The facts are like so:

1) There are probably around 5 loci of major effect that explain skin color variation in human beings (older pedigree analysis as well as new genomic data seems to be focusing on a number in this range).

2) One of the major loci, MC1R, is highly polymorphic in Europeans, constrained in dark-skinned populations to the ancestral consensus sequence (Africans, to a lesser extent in South Asians), and being driven toward fixation on a derived allele in East Asians.

3) A new locus in Europeans seems fixed and is responsible for the skin color depigmentation, but in both East Asians and Africans this locus is in an ancestral condition.

From these data I hypothesized that East Asians and Europeans were genetically isolated so they reached the same fitness optimums (assuming light skin has selective benefits at high latitudes) via different genetic architectures. In Europeans MC1R is diversified (either through frequency dependent selection or relaxation) because SLC24A5 handles the phenotype. In East Asians SLC24A5 is in an ancestral state, so MC1R has been coopted via the derived allele to generate the same end via different means.

I think I was wrong, and the evil Finns are to blame…they exhibit a rather high frequency of the MC1R allele which confers light skin in East Asians, but they’re in Europe! The idea that I threw out there is really contingent on powerful genetic barriers, beneficial alleles have a way of sweeping everywhere that they confer fitness all things being equal. Lactose tolerance is spread throughout Eurasia, and not because of massive population movements, while various malarial resistance alleles are pretty common across regions just where malaria is endemic. Perhaps there was a massive ice sheet in the middle of Eurasia, and Finns are the byproduct of an admixture event, but my understanding is that MC1R has been under positive selection for around ~10,000 years in East Asians. That’s pretty recent, and I don’t think that the genetic barriers across Eurasia were that strong for that period, they should been able to “borrow” SLC24A5 instead of throwing up their own trick de novo. Other shit is going down…and I’m sure that pleiotropy has its dirty hand involved here.

But I was wrong, though I did give it a good college try :)

Labels: ,


  1. I don’t get why you think you have proven yourself wrong. All that we would need for you to have been right, I think, is for the new SLC24A5 to have a head start in Europe, and the new MC1R a head start in Asia. Once these variants are established in their territories, other variants provide no fitness benefit. The Finn data would then be the result of neutral diffusion through contact.

  2. Interesting that Finns, who had contact with European depigmentation genes and Asian depigmentation genes, exhibit some of the highest overall depigmentation (skin and hair) in the world.

  3. I think I was wrong, and the evil Finns are to blame… 
    Mwa-ha-haa! We’re eeevil! 
    Where’d you get 10,000 years for MC1R selection? 
    The MC1R article said: 
    “Perhaps MC1R was the first gene to respond to selection for lighter pigmentation when modern humans moved into latitudes away from the equator, simply because, at this gene, any of a large number of knockout mutants will suffice. (…) Subsequently, a more effective mutation in another gene may have arisen, and, in sweeping through the population, it may have swamped any fitness differentials introduced by MC1R variation.” 
    Does SLC24A5 fit this hypothesis, or is the variant allele too the wrong age?

  4. I vaguely recall that the Finns speak a non-Indo-European language, so they probably have a chunk of Asian (Altaic?) ancestry not too long ago.  
    I wonder if the Hungarians (also with some Asian ancestry) have the same skin colour genes?

  5. Altaic 
    nnnnoooo!!!! jaakkeli, set him right!

  6. AAARGH! Sure. No, Finns do not speak an “Altaic” language and no, Finnish is not related to them. The family is called Uralic and has been spoken in Europe for as long as anyone knows – if you’re counting languages as “Asian” or “European” according to where you find the majority of Old World speakers or where the center of the range is, Indo-European languages are Asian and Uralic languages are European (being then the only remaining European language family!). Indo-European languages aren’t from northern or western Europe, so there is no mystery whatsoever in Europeans speaking non-IE languages, as most Europeans used to speak non-IE languages and have recently switched. It’s the IE speakers who’ve gone through a recent change! 
    Of course, in terms of geographical range, Uralic languages have only been known to have been spoken in the border region northern Asian and European population – Uralic languages are spoken mostly by European-looking peoples with some occasional Asian featurs and some Asian-looking peoples with partly European features. This is all very unsurprising and I can’t comprehend why it keeps baffling people and getting them to uncritically believe these crackpot ideas. It’s not that long from Finland to Siberia. No one gets baffled over why Greeks have some features in common with Lebanese and you don’t see people fantasizing about how these commonalities must mean that Greeks were born as a mix of swarthy Asians and blonde Germans. Well, OK, you do see that, but everyone with 2 braincells recognizes those fantasizers as crackpots. Why is it different for northerners?

  7. Why is it different for northerners? 
    the blood must be kept pure!!!

  8. I looked at those MC1R articles some more – it seems that Arg163Gln dominates in the Americas as well, although the sample sizes are awfully small. But it seems the allele can’t be younger than ~30000 years? 
    If the Paleolithic populations in Europe and East Asia populations were not as well connected before and around the LGM, the “different solutions to the same problem” model could work. 
    If Arg163Gln really is a selected variant, maybe there’s been more gene flow from slightly more southern populations to East Asia, resulting in that it has not become fixed there yet? 
    Wouldn’t it be interesting to see which alleles of the known skin colour genes can be found in the Ainu?