Elhaik and Ashkenazi ancestry: Scholarly infatuation or unhealthy obsession?

[Agamemnon]

This is meant to be a continuation of our conversation in the New DNA Papers thread.

In light of the recent media coverage of Das et al. 2016's peer-reviewed paper on the origins of Ashkenazi Jewry (which I've already mentioned here), one finds himself compelled to ask a very specific question: Considering Behar et al. 2013's conclusive rebuttal of Eran Elhaik's (who is one of the authors of the aforementioned study) former paper on this issue, isn't the inquiry simply pointless? And despite all the media hype around Elhaik's theories, isn't he basically arguing against the data?

By all means, discuss.

[Anabasis]

Well i will not advocate the Elhaik bu actual i realy hard to understand why he is reacted so much. While most of the east medeterenians close to shepardhaid jews admixtures, why not Ashkenazi Jews couldnt have Hazar ancestry?

Here is the link of the his academic paper. http://gbe.oxfordjournals.org/conten...vw046.full.pdf

Actually the model in 38th page probably disturbing some traditional jewish thesis. Could some one converted to Judaism? Yes why not!!! I will not suprise if his theories are proved by some other scientist. Or he is disproved by scientific way

[Shaikorth]

Well i will not advocate the Elhaik bu actual i realy hard to understand why he is reacted so much. While most of the east medeterenians close to shepardhaid jews admixtures, why not Ashkenazi Jews couldnt have Hazar ancestry?

Here is the link of the his academic paper. http://gbe.oxfordjournals.org/conten...vw046.full.pdf

Actually the model in 38th page probably disturbing some traditional jewish thesis. Could some one converted to Judaism? Yes why not!!! I will not suprise if his theories are proved by some other scientist. Or he is disproved by scientific way

IBD analysis and PCA analysis will put Ashkenazi Jews right next to Sephardic Jews. Elhaik's method of running people through Geno 2 (not the next gen but the old version) and then GPS'ing them puts Ashkenazis into Anatolia, but it would do the same for Sephardic Jews and even some South Italians. No one can seriously argue Sephardic Jews and Calabrians/Sicilians come from an "Anatolian Irano-Slavic mixture".

Elhaik probably knows this too, which is why he did not include Sephardics in the study. I have to say it looks like there is an intent to separate Ashkenazi from the "genetic jewish" group by any means necessary involved.

[Silesian]

Values for CHG for various groups in question.
https://drive.google.com/file/d/0B-O...w?pref=2&pli=1

CHG write up.
http://www.cam.ac.uk/research/news/f...ted-by-ice-age

[warwick]

This is meant to be a continuation of our conversation in the New DNA Papers thread.

In light of the recent media coverage of Das et al. 2016's peer-reviewed paper on the origins of Ashkenazi Jewry (which I've already mentioned here), one finds himself compelled to ask a very specific question: Considering Behar et al. 2013's conclusive rebuttal of Eran Elhaik's (who is one of the authors of the aforementioned study) former paper on this issue, isn't the inquiry simply pointless? And despite all the media hype around Elhaik's theories, isn't he basically arguing against the data?

By all means, discuss.

In principle, it is useful for someone to be out there advocating a radical theory. This researcher seems to have adopted the model of consistently advocating nonsense rather than plausible science. Serious linguists don't attach any credibility to these ideas.

Ultimately, by advocating nonsense, he attracts attention to himself. Perhaps that is the goal rather than serious inquiry.

[AJL]

I think it is quite obvious to anyone who has run enough genomes of Ashkenazi people through Gedmatch oracles or in utilities in R that the best non-Ashkenazi matches approximate overwhelmingly to nearby Jewish populations or other Levantine populations with a slight Central European pull. For example, one Ashkenazi kit's best non-Ashkenazi fits:

Eurogenes K13
Algerian_Jewish + Lebanese_Druze + Serbian + Tunisian_Jewish @ 2.266099

Dodecad V3
Morocco_Jews + Sephardic_Jews + Syrians + Slovenian @ 2.114876

puntDNALK12Modern
German_South + Turkish_Aydin + Turkish_Jew + Yemenite_Jew @ 2.589072

[Agamemnon]

In principle, it is useful for someone to be out there advocating a radical theory. This researcher seems to have adopted the model of consistently advocating nonsense rather than plausible science. Serious linguists don't attach any credibility to these ideas.

Ultimately, by advocating nonsense, he attracts attention to himself. Perhaps that is the goal rather than serious inquiry.

Over the years I've also assumed that "There is no such thing as bad publicity" is Elhaik's personal motto, on the other hand there's no level he isn't willing to stoop to, I mean associationg with a clown like Paul Wexler is about as low as you can possibly get.

[Agamemnon]

I think it is quite obvious to anyone who has run enough genomes of Ashkenazi people through Gedmatch oracles or in utilities in R that the best non-Ashkenazi matches approximate overwhelmingly to nearby Jewish populations or other Levantine populations with a slight Central European pull. For example, one Ashkenazi kit's best non-Ashkenazi fits:

Eurogenes K13
Algerian_Jewish + Lebanese_Druze + Serbian + Tunisian_Jewish @ 2.266099

Dodecad V3
Morocco_Jews + Sephardic_Jews + Syrians + Slovenian @ 2.114876

puntDNALK12Modern
German_South + Turkish_Aydin + Turkish_Jew + Yemenite_Jew @ 2.589072

Indeed, the sheer similarity between Sephardic and AJs is conveniently and consistently ignored by Elhaik, on the other hand I'm not really seeing that Central European pull you speak of, in fact the AJs' European admixture seems to be similar to that of Italian, Romaniote and French Jews by and large.

Take my dad's results for instance, the only test where he gets German in his 4-way oracle's non-Ashkenazi fits is Dodecad V3, but even there we have at least one fit which follows the general pattern I'm talking about, that is to say Southern European/Mediterranean admixture:


Dodecad V3
Portuguese + Syrians + Turkish + Tuscan @ 1.780126


Dodecad K12b
Bulgarians + Cypriots + Iraq_Jews + Murcia @ 1.475580


Dodecad K7b
O_Italian + Samaritians + Turkish + Valencia @ 0.778724


Eurogenes V2 K15
Greek_Thessaly + Kurdish_Jewish + Libyan_Jewish + Spanish_Galicia @ 4.570015


puntDNAK12Modern
French + Sicilian_West + Tunisian_Jew + Turkish_Aydin @ 0.878017


Harappa World
french + lebanese + lebanese-christian + tuscan @ 1.337348


MDLP K13
France + Italy_North + Lebanese + Syrian_Jew @ 0.000000
Cretan + North_Greek + Spain + Syrian @ 0.273508
Italy-Friul + Occitan + Samaritian + Syrian_Jew @ 0.287489


^^The Southern European admixture is highlighted in red, I added the other MDLP K13 fits to highlight what "French" or "France" actually means here (since France is quite diverse and regionally stratified, here this undeniably stands for the southernmost French hence the "Occitan" in the last fit). Could've done the same for most of the other tests, there's no shortage of such fits in general.

[AJL]

Indeed, the sheer similarity between Sephardic and AJs is conveniently and consistently ignored by Elhaik, on the other hand I'm not really seeing that Central European pull you speak of, in fact the AJs' European admixture seems to be similar to that of Italian, Romaniote and French Jews by and large.

I suspect that the direction of the pull away from the Near East will differ from case to case, perhaps dictated by distant admixture, or by drift, or by chance of recombination.

Another AJ kit, with the same oracles, gets a much more Balkan pull but perhaps suggesting distant Sephardi ancestry:

Eurogenes K13
Bulgarian + Greek_Thessaly + Italian_Jewish + Lebanese_Druze @ 3.796003

Dodecad V3
Balkans + Druze + Georgia_Jews + Portuguese @ 1.452617

puntDNALK12Modern
Hungarian + Kurdish + Libyan_Jew + Libyan_Jew @ 1.247540

In this last case, if you leave out "Libyan Jew" and average Hungarian and Kurdish, you should find a spot somewhere in Turkey.

[Agamemnon]

I suspect that the direction of the pull away from the Near East will differ from case to case, perhaps dictated by distant admixture, or by drift, or by chance of recombination.

Another AJ kit, with the same oracles, gets a much more Balkan pull but perhaps suggesting distant Sephardi ancestry:

Eurogenes K13
Bulgarian + Greek_Thessaly + Italian_Jewish + Lebanese_Druze @ 3.796003

Dodecad V3
Balkans + Druze + Georgia_Jews + Portuguese @ 1.452617

puntDNALK12Modern
Hungarian + Kurdish + Libyan_Jew + Libyan_Jew @ 1.247540

In this last case, if you leave out "Libyan Jew" and average Hungarian and Kurdish, you should find a spot somewhere in Turkey.

That Eurogenes K13 fit you just posted is quite telling, as you probably know I suspect that for the most part the European ancestry of Western Jews (Ashkenazi, Sephardic, Maghrebi etc.) is Greek and Aegean in origin, so the presence of "Greek_Thessaly" is intriguing as it shows up more than once, not only in AJs but also in Sephardim and other Western Jews.

Although my dad has no non-Ashkenazi fits in K13, he stills gets "Greek_Thessaly" in 3 of the 4-way oracle fits:

Ashkenazi + Greek_Thessaly + Italian_Jewish + Italian_Jewish @ 2.938966
Ashkenazi + Greek_Thessaly + Italian_Jewish + Sephardic_Jewish @ 3.182374
Ashkenazi + Ashkenazi + Greek_Thessaly + Italian_Jewish @ 3.357651