search this blog

Thursday, February 27, 2014

Khazar shmazar

Human Biology recently posted several open access manuscripts dealing with the topic of Jewish origins (see submissions from 2013 here). One of these preprints is essentially a rebuttal to an Eran Elhaik paper from a couple of years ago, which argued that a substantial part of Ashkenazi Jewish ancestry was derived from within the Khazar Empire. The leading author of the new preprint is Doron M. Behar, but thirty people in all, many of them well known scientists, have put their names on it. Here's the abstract:

The origin and history of the Ashkenazi Jewish population have long been of great interest, and advances in high-throughput genetic analysis have recently provided a new approach for investigating these topics. We and others have argued on the basis of genome-wide data that the Ashkenazi Jewish population derives its ancestry from a combination of sources tracing to both Europe and the Middle East. It has been claimed, however, through a reanalysis of some of our data, that a large part of the ancestry of the Ashkenazi population originates with the Khazars, a Turkic-speaking group that lived to the north of the Caucasus region ~1,000 years ago. Because the Khazar population has left no obvious modern descendants that could enable a clear test for a contribution to Ashkenazi Jewish ancestry, the Khazar hypothesis has been difficult to examine using genetics. Furthermore, because only limited genetic data have been available from the Caucasus region, and because these data have been concentrated in populations that are genetically close to populations from the Middle East, the attribution of any signal of Ashkenazi-Caucasus genetic similarity to Khazar ancestry rather than shared ancestral Middle Eastern ancestry has been problematic. Here, through integration of genotypes on newly collected samples with data from several of our past studies, we have assembled the largest data set available to date for assessment of Ashkenazi Jewish genetic origins. This data set contains genome-wide single-nucleotide polymorphisms in 1,774 samples from 106 Jewish and non- Jewish populations that span the possible regions of potential Ashkenazi ancestry: Europe, the Middle East, and the region historically associated with the Khazar Khaganate. The data set includes 261 samples from 15 populations from the Caucasus region and the region directly to its north, samples that have not previously been included alongside Ashkenazi Jewish samples in genomic studies. Employing a variety of standard techniques for the analysis of populationgenetic structure, we find that Ashkenazi Jews share the greatest genetic ancestry with other Jewish populations, and among non-Jewish populations, with groups from Europe and the Middle East. No particular similarity of Ashkenazi Jews with populations from the Caucasus is evident, particularly with the populations that most closely represent the Khazar region. Thus, analysis of Ashkenazi Jews together with a large sample from the region of the Khazar Khaganate corroborates the earlier results that Ashkenazi Jews derive their ancestry primarily from populations of the Middle East and Europe, that they possess considerable shared ancestry with other Jewish populations, and that there is no indication of a significant genetic contribution either from within or from north of the Caucasus region.

I'm really not sure what to make of all of this attention that the Khazar hypothesis is still getting? It's been obvious for a while now that in terms of genetic structure Ashkenazi Jews are basically a group of East Mediterranean origin. But Elhaik's paper did get a fair bit of media coverage, so I suppose after that a rebuttal was to be expected.

In any case, I'm not complaining. This paper includes a very interesting genotype dataset of many previously unpublished samples, which I tested last week with PCA (see here).


Behar, Doron M.; Metspalu, Mait; Baran, Yael; Kopelman, Naama M.; Yunusbayev, Bayazit; Gladstein, Ariella; Tzur, Shay; Sahakyan, Havhannes; Bahmanimehr, Ardeshir; Yepiskoposyan, Levon; Tambets, Kristiina; Khusnutdinova, Elza K.; Kusniarevich, Aljona; Balanovsky, Oleg; Balanovsky, Elena; Kovacevic, Lejla; Marjanovic, Damir; Mihailov, Evelin; Kouvatsi, Anastasia; Traintaphyllidis, Costas; King, Roy J.; Semino, Ornella; Torroni, Anotonio; Hammer, Michael F.; Metspalu, Ene; Skorecki, Karl; Rosset, Saharon; Halperin, Eran; Villems, Richard; and Rosenberg, Noah A., No Evidence from Genome-Wide Data of a Khazar Origin for the Ashkenazi Jews (2013). Human Biology Open Access Pre-Prints. Paper 41.

Elhaik E. The missing link of Jewish European Ancestry: contrasting the Rhineland and Khazarian hypotheses. Genome Biol Evol. 2012. doi:10.1093/gbe/evs119, Advance Access publication December 14, 2012.

See also...

Near Eastern origin of Ashkenazi Levite R1a


Helgenes50 said...

Do you have an idea of what caused the founder effect in the Ashkenazi population ?

In the oracles, Ashkenazi often are close to the Italians or Southern Italians.

It's why we often read that they are descendants of Romans converted.
But is it not rather because they share a same East Mediterranean origin,
an origin much older

Davidski said...

This isn't really my area of interest, but persecution and shunning of Jews by European gentiles, including pogroms, are often mentioned in this context.

That might be true in part, but basically what I think happened is that only small groups of Jews made it deep into Europe from the Mediterranean region, and these groups then isolated themselves from the locals because of religious and cultural differences. Also, most Europeans at the time were serfs, and marrying a serf wasn't cool.

Helgenes50 said...

Thanks for your reply

Ponto said...

PCAs mostly use two dimensions. Using three dimensions the Jews will separate cleanly from Southern Italians. The researchers themselves mention not to put too much into the location of the ethnic groups on the PCA. PCAs of Europeans usually show their geographic location with some exceptions namely the Basques and Sardinians who are far removed to the west, Croatians who are located near the Hungarians and away from the Mediterranean, and Jews who end up floating between Greece and Italy and Africa.

Speaking personally, Jews are not particularly close to me when using IBD segments. Greeks, Tuscans, Bulgarians and Sardinians are closer to me than any Jews, and the first West Asians that are closest on IBD are Cypriots.

I would not rely on Oracle as it is wildly inaccurate, actually useless for Central and SE Mediterranean Europeans. Totally useless.

Davidski said...

Ashkenazi Jews pull away from Southeast Europeans on PCA in the lesser dimensions because they're basically like an extended family. But they really don't differ very much from Southeast Europeans and even most other East Mediterranean groups in terms of overall genetic structure. So it's kind of like running a bunch of Italian second cousins on PCA, seeing that they cluster together and apart from everyone else, and then saying that they differ genetically from other Italians. Well yes they do, but not really.

Sgt said...

1. Endogamy was imposed by outside forces. Wherever the Caliphate became established or a nation adopted the Nicene creed intermarriage with Jews became a capital offence. Lithuania did not full become fully Christian until the 15th Century which may explain why I see longer IBD lengths between AJ and Lithuanians compared to Middle-Easterners or Tuscans {recency}.

2. Admixture. By your own calculators European Jews are ~35% East-Med, Mid-East which may be the largest concentration but still a minority. PCA results are not conclusive: the child of a Norwegian and a Saudi does not become Italian because that's where they might sort. Formal tests of Admixture between AJ and even other AJ or Sephardic, let alone others, becomes difficult as inconsistent LD decay rates and other features of founder effects and drift come into play. Yet,I still have found positive Alder {albeit at a low Z} between AJ and Chuvash, Russians, North Ossetians, Kumyks, Adygei, Lezgins, Kurds, Pathans and Tadjiks. Given concentrations of S. Central Asian, North Caucasus and E Euro admixture among many AJ {from potentially "Khazar-type" peoples} this could account from between 10 and 20 per cent of AJ ethnogenesis. Maybe it's a higher percentage, but as of yet we have no tools to peel these layers back.

One needs to remove politics and rancor from these published studies!

Davidski said...

Yeah, OK, but Ashkenazi Jews can't be compared to someone who's Norwegian/Saudi. They do cluster with several other Jewish groups on different PCA and across different dimensions, and this isn't just a happy coincidence.

Also, those positive East European and North Caucasus-like signals you're seeing aren't really surprising, because Ashkenazi Jews do seem to have some ancient Iranian and, as you note, more recent East European ancestry. For instance, Ashkenazi Levites carry an Iranian-specific R1a subclade.

10 to 20 per cent of this Khazar-like ancestry sounds about right, although I'd say it's closer to 10 for most. Nevertheless, Khazar-like still isn't Khazar.

Sgt said...

Maybe; but nobody is telling the Hungarians all they have from the Magyars is language and ~3% of the genetics :) -- Point being why so much intensity to the issue of AJ ancestry?

Davidski said...

I honestly don't care. I'm as objective on this issue as anyone you'll ever find.

Khazar shmazar, I say.

About Time said...

The Khazar hypothesis is popular because it appeals to people who want to deny the Jewishness of European Jews. The biggest proponents tend to be in the Muslim world who can't believe "those European people" could possibly be the real McCoy after all this time.

There was a real Khazar state. It was multiethnic and not homogeneous. Probably some people fleeing Islamic conquests in the Mideast and C Asia went there, among others. There were lots of Jewish communities in the Arab world before that.

Persecution was SOP in lots of places. Educated people/communities aren't always appreciated.

Every Muslim knows about this, but it's some kind of blank spot for western history. But look at dates. Even if some European Jews lived in Khazaria, that doesn't mean they weren't descended from the original Tribe/Kingdom of Judah (technically plus tribes of Benjamin and Levi - other tribes went AWOL long ago).

About Time said...

Since people never do their own homework, here are dates:

Battle of Khaybar 629 AD
Islamic conquest of Persia 633-654 AD
Khazar state 618?-648? AD

Conceivably some people went to a Khazaria earlier because things were going sour long before Khaybar (which again was a major turning point in Islamic conquests). After Khaybar, even the dummies who failed to see the writing on the wall went north.

Khazaria traded with Venetians in the Crimea and Varangians and some others who were around. North Europe was still pagan back then (as in, real pagans who were not so nice). Read Ibn Fadlan for first hand descriptions.

Lloyd said...

Dear "Davidski",

I am an FTDNA client who just discovered your various Eurogenes programs on gedmatch. I was a bit skeptical at first from simply reading the admixture results, but I was amazed to discover your Oracle functions, which look stunningly precise and relevant. Plus, I can see by reading this blog that you are incredibly knowledgeable about human genetics.
I do have a few questions though, about my results and their interpretation and meaning. Would you please be willing to answer them, whether here or by email or elsewhere? I'm sorry that this message isn't related to the subject under which it is posted, but I didn't know where else to contact you. I understand you must be quite busy but I'd be really happy if could you shed some light on a few points.

Best wishes.

Davidski said...


Please post your questions at the link below, and I'll reply later today. Thanks

אחיקם גדליהו ישעיהו בן הלל said...

The Khazar hypothesis is currently dying a horrible death.

One of the best arguments in Behar et al's pre-print is that if were were to take Armenians & Georgians as a surrogate Khazar population (which makes no sense, but for argument's sense we'll try), the Iranian-Iraqi-Georgian Jewish cluster obviously is more "Khazar" than the Sephardi-Ashkenazi cluster.

As Polako said, Ashkenazi Jews are a typically East Mediterranean-North Levantine group clustering between Greeks & Cypriots on PCA plots.

I tend to be skeptical of the Roman-convert hypothesis, one of the reasons for my skeptical approach is that there isn't much IBD sharing with Italians and that pre-exilic Judeans might've been more similar to Cypriots than to modern-day Levantines.

Genome-wide studies of ancient Israelite remains will tell, but it looks like we'll still be splitting hairs in order to quantify the degree of "European" admixture.

About Time said...

The problem is Judea was a kingdom since at least 900s bc, so before any other putative pop that is compared with (Italians, Arabs, and N Eurppeans). Their ethnogenesis was later, after breakup of Israel/Judah as far as anyone knows.

Others like ancient Assyria are long gone, so who knows if any modern pop really represents them. Best clues might be in places beyond the reach of easy admixture, where odd components show up (like Kalash/Orkney).

Also, there is no reason to assume these pops were ever homogeneous, so it's like comparing a palimpset to an unknown assemblage of pops with a temporarily unified culture/religion.

Judea was the poor cousin of the bigger Omride kingdom (according to Finkelstein's minimalism) which was probably more heterogeneous. We get skewed picture of ancient Near East by thinking of the more low key culture that led to Judaism. The flashier cultures like Assyrians didn't last.

Egyptians are still around in some form, but others like Edom, Moab, etc are long gone. Really ancient genomes are the only comparables.

V Robazza said...

Dnatribes state AJ origins bordering on Czech, polish and slovak areas.

But then its too far from SJews

Grey said...

The Khazar hypothesis never made sense as the Jews in the Crimea were allies of the Turkic tribes like the Khazars so after the destruction of Byzantium by the Turkic tribes the obvious direction for Jews to go was *towards* the new Ottoman state not the completely opposite direction.

Kapak said...

"Maybe; but nobody is telling the Hungarians all they have from the Magyars is language and ~3% of the genetics :) -- Point being why so much intensity to the issue of AJ ancestry?"

Not that I disagree with the point you're trying to get across, I just have to say this regarding Hungarians:

Hungarians are 1% East-Asian/Siberian..... Ancient Magyars and various Turkic intruders in east Europe would have been, 10% East-Asian/Siberian... we can safely say the Cumans were around this % based on things like Baibars the blonde mameluke and what not..... So why not Magyars, Pechenegs, etc. as well? A Hungarian is 10% Magyar+Cuman+Hunnic mix. The rest is mainly Slavic people with some Germanified Celtic people.

Sgt said...

@ Kapak:
I haven't seen many recent papers about Hungarians but there are two by Kovácsné Csa´nyi; one in 2008 and a Phd thesis 2009. Interesting but outdated by now, I'm sure. One needs to take all of these estimates with a grain of salt. But you're right for the most part Hungarians seem to be Central-East Euro with some Med; However the Magyar spirit lives on -- brother tribes to the Khazar-Shmazars :)

barakobama said...

Who likes my new profile picture? If it wasn't for him I wouldn't have Y DNA R1b L11(probably Df27).

Grey said...

nice hat