November 02, 2012

ALDER estimates of East Eurasian admixture in Europe

I used the 1-reference method of ALDER to infer lower bounds of East Eurasian admixture in a few European populations. This method does not include a statistical test of admixture (as does the 2-reference one or the f3 test), but we can probably reasonably suppose that some such admixture did take place on the combined evidence of the f3 test and ADMIXTURE evidence.

In any case, I took the East Asian populations of Loh et al. (2012) which had no evidence of admixture with either ALDER or the f3 test, and also a few populations from Rasmussen et al. (2010) that included representatives of Siberian Uralic speakers, as well as the three main branches of narrow-sense Altaic (Turkic, Mongolian, Tungusic), and estimated lower bounds of admixture for a set of European populations. Results can be seen below:

The evidence for admixture appears most convincing in the 1000 Genomes Finns and HGDP Russians where the +/- interval does not intersect or approach zero irrespective of the Asian population chosen. For these populations, the percentages vary from ~4-5% for the "pure" East Eurasians to ~10% for some Siberian groups such as Selkup and Altai. Thes latter carry some West Eurasian admixture, so it makes sense that a greater deal of admixture with them is necessary to account for the observed "East Eurasian" influence. And, indeed, it is probably via such "intermediate" Siberian populations that some East Eurasian ancestry flowed into Europe, rather than via the relatively untouched populations of the Far East.

PS: Note that this probably represents the most recent signal of admixture, and not the older and more general "North Eurasian"/Amerindian-like admixture that, as Loh et al. mention in their paper cannot be captured with ALDER.


M. Myllylä said...

How do you have interpret the gene flow between Easts Europeans and Asians? How we can interpret it when we know that Southern Europeans dont own all ancient European genes, thus cannot be used as a discriminator.

Dienekes said...

South Europeans do not feature in this particular analysis.

M. Myllylä said...

I meant how we know the origin of these transmitted genes and the direction of the gene flow. You use quotes with the word "pure" indicating that the gene flow direction id known. I myself am not able to interpret it.

Dienekes said...

I meant how we know the origin of these transmitted genes and the direction of the gene flow.

Both the f3-statistic and ALDER test for admixture using different genetic mechanisms (allele frequency correlations and linkage disequilibrium). So, we can be fairly sure that Finns and Russians have East Eurasian admixture. Some Siberian populations also have European admixture. Genes went both ways.

M. Myllylä said...

This was not unclear for me. More it was the possibility for bidirectional gene flow in contacts between two populations which leads to common ancestry for both. As far as I can understand the unidirectional gene flow is possible only in certain conditions, in case of certain migration types.