Via Yan Klimentidis' blog.
Yuval Itan et al. A worldwide correlation of lactase persistence phenotype and genotypes. BMC Evolutionary Biology, 2010. Open access PDF.
The authors explore the differences between actual lactase persistence phenotypes, i.e. the real apportions of adults who can digest milk, and the presence of alleles thought to cause this phenotype. They find that in many parts of the world, notably SE Europe, West Asia and most of Africa, the number of actual milk drinkers is often very different than known genetics could predict.
Some striking examples:
- Central Italians: actual LP: 82% - predicted by genetics: 21%
- Afghan Tajiks: actual LP: 18% - predicted: 51%
- Jordanians: actual LP: 76% - predicted: 20%
- Balochis: actual LP: 38% - predicted: 0%
- Wolof (Senegal): actual LP: 51% - predicted: 0%
- Sudanese: actual LP: 88% - predicted: 45%
Actual lactase persistence (phenotype):
Predicted lactase persistence (known alleles):
Predicted lactase persistence (known alleles):
Genetic canalisation?; distinct genotypes producing the same phenotype.
ReplyDeleteI call that "evolutionary convergence" but the same idea.
ReplyDeleteHello.
ReplyDeleteYou wrote : "Afghan Tajiks: actual LP: 18% - predicted: 51%"
Are you sure it's not the reverse? 51/18 rather than 18/51
Yes, that's what the paper says (you can check yourself: table 1). It's not the only case where genetic predictions are larger than actual phenotype frequency but it is one with the largest difference. Another quite striking one are hungarians (predicted: 86%, actual 56%). There are a few others (Iranians, Indians, Irish, North Italians) but the differences are smaller.
ReplyDeleteThank you. Interesting.
ReplyDeleteA visitor to Mauritania recently commented to me on its endless flat sandy terrain and the goats. LP there would of necessity be derived from their milk.
ReplyDeleteMauritania scores rather high in actual LP phenotype (although the point samples are from the Wolof, not the Arabs) but the alleles involved are a mystery to date. The closest Arab sample is in Tunisia and it scores very low instead.
ReplyDelete