Wednesday, August 8, 2012
Khazar origins of European Jews finally confirmed?
This week saw the release of yet another study on the origins of European Jews, titled "The Missing Link of Jewish European Ancestry: Contrasting the Rhineland and the Khazarian Hypotheses". The author seems quite content that he's demonstrated an unambiguous genetic link between European Jews and the Khazar Empire. But has he really?
I do like the analysis presented is this paper. It's varied, thorough, and looks at Central European Jews separately from Eastern European Jews, with some interesting outcomes. But in the end, I think the author fumbled his interpretation of the results. His mistake was treating the Armenian reference sample as a Caucasus group and a proxy for the gene pool of the Khazar Empire. Thus, when the Jewish samples showed strong affinity to the Armenians, the author mistook this as a signal of Khazar ancestry in Jews, because the Khazar Empire included parts of the Caucasus.
But what do modern Armenians of the South Caucasus have to do with ancient Khazars of the Pontic Caspian Steppe? Not much, I'd say. Armenians aren't even a useful Caucasian reference set in my opinion. They're better treated as an Eastern Anatolian group, due to their high affinity to Mediterranean and Middle Eastern populations.
Moreover, they show low North/East European genetic ancestry, and very little East Eurasian admixture, which is actually the sort of stuff we'd want in a proxy for the largely Turkic inhabitants of the Khazar Empire in what is now Southern Russia. Based on my own analyses of Jewish genomes, I'd say that Ashkenazi Jews are genetically an Eastern Mediterranean group, but with various other admixtures, like Western European, Eastern European, Eastern Anatolian, and even African and East Asian. Does that mean ancestry from the Khazar Empire? Perhaps in part, but it's hard to say for sure.
So, what could be a sure signal of Khazar influence in modern Jews? The best bet is probably Y-chromosome haplogroup R1a-Z94, which reaches a high frequency among Ashkenazi Levites. This marker is also very common among modern Indo-Iranian and Turkic groups, so it's not difficult to imagine its presence in ancient Khazaria. The only problem is that it's also found among Arabs, who obviously share deep Semitic roots with the Jews. That's why it's not possible to say at the moment if the Jewish R1a-Z94 is of Semitic, Khazar or some other origin, like, for example, Persian. Someone should look into that.
By the way, as per the supervised ADMIXTURE bar graph from the study, it's interesting to see the much higher levels of “Eastern European” influence in Eastern European Jews than in Central European Jews. I suppose that’s to be expected, considering the geography, but the reason I find it interesting is that it indicates recent Slavic introgression into the Eastern European Jewish gene pool. That's because if it was older, like from Khazaria, then Central European Jews would also carry it at decent levels.
Eran Elhaik, The Missing Link of Jewish European Ancestry: Contrasting the Rhineland and the Khazarian Hypotheses, Submitted on 6 Aug 2012, arXiv:1208.1092v1 [q-bio.PE]
Update 22/01/2013: The full study is now out at Oxford Journals as an open access article. See here.
Update 07/09/2013: A presentation titled "No indication of Khazar genetic ancestry among Ashkenazi Jews" will be held at the ASHG 2013 annual meeting in October. This appears to be a rebuttal to Elhaik's claims. For more info on that see here.
Most Ashkenazi maternal lineages are of European origin