Ashkenazi Genetics Illuminate History

[tomcat]

http://www.sciencedaily.com/releases/2013/01/130116195333.htm

[mollyblum]

Very interesting. But I guess I am a little confused? How is it that most of us with Ashkenazi Jewish ancestry show up in PF with Middle Eastern percentages? If this study were true shouldn't my ancestry show up as South Asian or whatever the Mongolian/Khazar population is considered?

Am I reading the article wrong?

AB

[vinnie]

There's a robust discussion of this paper at 23; Davidski's two quotes are below. Another poster noted that the study is being posted on neo-Nazi and other anti-Israel/Jewish websites.

"There's a major problem with this study, because it claims there's high IBD sharing between "Caucasus" groups and Ashkenazi Jews. But this hasn't been reproduced by anyone else.

I think what happened is that this IBD sharing reflects some sort of Mediterranean links between Armenians and Jews, or the author failed to filter out Caucasus samples with significant Russian and Ukrainian ancestry, which is also present in many Ashkenazi Jews.

I know for a fact that many of the public Caucasus samples carry such recent Russian and Ukrainian ancestry, because I have them as part of my dataset, and this could easily skew the results of a study such as this."

"... It's really not a bad analysis overall. I'm just not sure about that IBD data, which is actually a key part of the evidence offered in support of the Khazar theory. It looks like an error to me, possibly due to a few outliers not being filtered out.

But whatever...it's obviously impossible to prove that Ashkenazi Jews don't have some Khazar ancestry, especially since they do cluster close to Anatolian Turks and Armenians. So anyone can make the claim based on any old genetic map, and it's hard to tell them they're definitely wrong.

I think it's one of those very specific things that can only be proven with extremely high resolution Y-DNA and full sequence mtDNA data. I have a hunch that finding the source of the Ashkenazi R1a-Z93 will be the clincher."

[MMaddi]

There's a robust discussion of this paper at 23; Davidski's two quotes are below. Another poster noted that the study is being posted on neo-Nazi and other anti-Israel/Jewish websites.

Yes, this is a problem with this theory, how it's used by anti-semitic types. They use the theory of "Khazar origin of European Jews" to say that Jews don't have any right to claim Israel. I'm not accusing the scientist behind the study of being anti-semitic, but you have to be careful of how your work may be used.

I'm not sure the study actually claims that the ultimate origin of European Jews is solely Khazar, but it seems to put more emphasis on that than any other origin. I think it's clear the European Jews surely owe a large part of their origin to the Levant. As mollyblum notes above, PF and other admixture analysis tests always show a large Mideast component among Ashkenazi Jews. In fact, I believe that there is a strong affinity between Ashkenazi Jews and Palestinians. If the thesis of this study is correct, does this mean that Palestinians' origin is also Khazar? Yet no one makes that sort of claim.

[josh w.]

Vinnie, I share your doubts. There is evidence of a Khazar or central Asian contribution to Jewish Y dna. However the contribution appears to affect only around 10% of the Ashkenazi population. The Yiddish language also reflects a contribution in the commonly used word 'Khazarai'.

On the other hand, most autosomal analyses including Behar's own views point to a European contribution. The usual finding is that there is a clear Italian contribution to the Ashkenazi population. The German contribution is less clear. First of all there is debate as to whether the Yiddish population came mainly from Bavaria or mainly from the Rhineland. Secondly, autosomal studies do not agree as to whether there is a central European contribution to the Ashkenazi population. None of the authors of the studies reported a central Asian component.

I am personally aware of the Jewish presence in the Caucasus. My Y dna line is clearly Ashkenazi but my closest match is an Ashkenazi from Georgia. This was probably the result of the forced migration of Russian Jews to southeastern Russia and the Caucasus in the 1800s. In terms of ancient history, I am a member of a Y dna Ashenazi J2a* cluster. The closest non Jewish J2a* cluster is from the Caucasus. My guess is that both lines migrated from the Zagros mountains where J2a originated, long before the Khazar era.

[ursus]

The one thing that I have learned from FF results is that although some Khazars may have intermarried into the the Eastern European Jewish population, it isn't the major source of European Jewry.

My parents were Jews who were born in Germany as were all of my ancestors back to my 3-great-grandparents, except for one who was from from the Alsace. Although I have not been able to trace all of my lines back further, I appear to be entirely of German and French Jewish ancestry for several centuries including some lines that can be traced to the Frankfurt ghetto in to the 14th century.

Despite that, I match 100s of people with Jewish ancestry from Eastern Europe (not that I have any idea what the actual relationships are. ). The logical explanation is that indicates that Eastern European Jews came from Germany. Granted that the migration could have been the other way around -- my ancestors could ahve come from Eastern Europe to Western Germany, but I have never seen that explanation of Western German Jewry given.

Then there is the fact that Eastern European Jews spoke Yiddish which is clearly basically German and supposedly a Rheinland dialect. Why would they have spoken that language if most of their ancestry wasn't from Germany?

[josh w.]

I hope the implications of my last point are clear. My Ashkenazi cluster and the Caucasian cluster are both part of the paragroup J2a*. At this point, both clusters show the same SNP pattern although that may change in the future regarding downstream SNPs. The degree of haplotype similarity suggests the lines separated thousands of years ago. It is not at all necessary to invoke the Khazars to explain the overlap.
In both the Ashkenazi and the southeast Russia-Caucasian populations one of the the most common Y haplogroups is J. It is likely that they also share autosomal SNPs. In the case of Y dna J there is no need to invoke the Khazars. The Tigris-Euphrates connection appears to be sufficient

[vinnie]

In fact, I believe that there is a strong affinity between Ashkenazi Jews and Palestinians. If the thesis of this study is correct, does this mean that Palestinians' origin is also Khazar? Yet no one makes that sort of claim.

Great point, Mike. I can't remember the name of the paper, but one came out a few years ago that showed AJs occupying a genetic space between S.Italians and Palestinians.

[MMaddi]

The one thing that I have learned from FF results is that although some Khazars may have intermarried into the the Eastern European Jewish population, it isn't the major source of European Jewry.

My parents were Jews who were born in Germany as were all of my ancestors back to my 3-great-grandparents, except for one who was from from the Alsace. Although I have not been able to trace all of my lines back further, I appear to be entirely of German and French Jewish ancestry for several centuries including some lines that can be traced to the Frankfurt ghetto in to the 14th century.

Despite that, I match 100s of people with Jewish ancestry from Eastern Europe (not that I have any idea what the actual relationships are. ). The logical explanation is that indicates that Eastern European Jews came from Germany. Granted that the migration could have been the other way around -- my ancestors could ahve come from Eastern Europe to Western Germany, but I have never seen that explanation of Western German Jewry given.

Then there is the fact that Eastern European Jews spoke Yiddish which is clearly basically German and supposedly a Rheinland dialect. Why would they have spoken that language if most of their ancestry wasn't from Germany?

You make some very good points here. While this study may make some genetic connections between Ashkenazi Jews and Khazars, notwithstanding the problems noted by Davidski (as quoted by vinnie above), your points are much more convincing to me.

Genetic evidence is part of the equation, but the linguistic evidence that you cite (along with known historical migrations) tips the scales to a Rhineland, not Khazar, origin for the bulk of Eastern European Jews. The fact that your French/German Jewish ancestry gives you FF matches to many Eastern European Jews is as strong or stronger genetic evidence than the Khazar study offers.

[MMaddi]

Great point, Mike. I can't remember the name of the paper, but one came out a few years ago that showed AJs occupying a genetic space between S.Italians and Palestinians.

Yes, every PCA plot of autosomal DNA admixture results that I know of shows similarities between Sicilians/S. Italians, Greeks, Cypriots and Ashkenazi Jews. An example of this is a PCA plot generated from Davidski's Eurogenes Project - http://img12.imageshack.us/img12/509...pedec82010.png.

Note the fairly tight cluster for the ethnic groups I mentioned. And see how Turkish and Caucasus populations are separate from the Ashkenazi, et al. cluster, although not far off. If the bulk of Ashkenazi Jews descend from the Khazars, you would expect them to cluster with the Turks, et al., who are geographically closer to the area where the Khazar Empire ruled, not with eastern Mediterranean populations, as they do. Also note that there are a couple of AJ outliers located between Germans and Romanians in the Eurogenes plot.

[JTR]

Can someone summerize the two theories? Please!

Also, here are my McDonald Results in light of these issue - would like your guys feedback on them.

Most likely fit is 53.9% (+- 7.3%) Europe (all Western Europe) and 26.1% (+- 4.4%) Europe (all Northeast Europe) which is 80.0% total Europe and 18.8% (+- 4.0%) Mideast (various subcontinents) and 1.3% (+- 0.8%) S. Asia (various subcontinents)

The following are possible population sets and their fractions, most likely at the top

English= 0.466 Lithuani= 0.284 Jewish= 0.226 Sindhi= 0.023 or
English= 0.470 Lithuani= 0.279 Jewish= 0.224 Pathan= 0.027 or
English= 0.448 Lithuani= 0.297 Jewish= 0.239 N_India= 0.016 or
English= 0.443 Lithuani= 0.302 Jewish= 0.243 S_India= 0.012 or
English= 0.608 Lithuani= 0.221 Armenian= 0.164 N_India= 0.007 or
English= 0.606 Lithuani= 0.222 Armenian= 0.165 S_India= 0.006 or
English= 0.612 Lithuani= 0.218 Armenian= 0.163 Sindhi= 0.007 or
English= 0.612 Lithuani= 0.217 Armenian= 0.163 Pathan= 0.008 or
English= 0.607 Lithuani= 0.223 Armenian= 0.170 Hazara= 0.000 or
French= 0.516 Lithuani= 0.347 Armenian= 0.117 N_India= 0.020

This is probably in fact ¾ German and the rest Jewish. But there could in fact be other northern or western European in there plus the Jewish.

[josh w.]

JTR, researchers such as Atzmon hold the more popular view that some Jews of ancient Israel migrated to Italy and the south coast of Europe during the Diaspora. With the rise of Christianity they migrated to Germany. Later many German Jews migrated to eastern Europe: Prague, Poland and points east. Autosomal research generally supports this view. A few studies have found both a Middle (Near) Eastern component and an Italian component in the Ashkenazi population.

The minority view is that eastern European Jews stem mainly from the Khazars a Turkic people who migrated to the Black Sea area. The debate is over whether the Khazar contribution is major or minor. Elhaik analyzed already collected autosomal data and found a pattern indicating that the Khazar contribution was major.

Most of the posts on this thread lean to the German migration explanation. In addition to my previous comments supporting this view I have located Ellhaik's paper and have noted questionable assumptions of his research. Elhaik assumed that modern Caucasian populations could serve as a proxy for the Khazars of 1000 years ago. On the basis of limited analyses he asserted that the Caucasian populations were genetically differerent from the Semitic populations of the Middle East. This assertion is completely contradicted by Yunusbayev's finding that the Caucasian populations are primarily of Middle Eastern origin. Elhaik also presumed that modern day Palestinians could serve as a proxy for ancient Israelites. While there is significant overlap between Jewish and Palestinian lines, Elhaik seems to be unaware of the Moslem Arabic migration into Palestine long after the Jews had left---e.g. as reflected in Y dna differences in J1/J2 ratios between Jews and Palestinians. On this issue it is not surprising that Jews are more like Caucasians than they are like Palestinians. Lebanese are also more like Caucasians and they certainly are Semitic.

[Javelin]

Elhaik assumed that modern Caucasian populations could serve as a proxy for the Khazars of 1000 years ago. On the basis of limited analyses he asserted that the Caucasian populations were genetically differerent from the Semitic populations of the Middle East. This assertion is completely contradicted by Yunusbayev's finding that the Caucasian populations are primarily of Middle Eastern origin.

The choice of Armenians as a proxy for Khazars was particularly dumbfounding.

First, Armenians are from the southern Caucasus, not too far from the northern Near East. A better choice might have been Tatars, Adyghei, Russians from the Volga area, or a similar population.

Second, some diaspora Armenians often have measurable descent from Near Eastern Christian populations who themselves share common ancient ancestry with Jews (e.g. Maronites, Assyrians, Copts). So depending on the exact nature of the Armenian sample, those Ashkenazim might be genetically similar to Armenians because those Aremnians have some Near Eastern influence, rather than the Ashkenazim necessarily having Caucasus influence.

[JTR]

JTR, researchers such as Atzmon hold the more popular view that some Jews of ancient Israel migrated to Italy and the south coast of Europe during the Diaspora. With the rise of Christianity they migrated to Germany. Later many German Jews migrated to eastern Europe: Prague, Poland and points east. Autosomal research generally supports this view. A few studies have found both a Middle (Near) Eastern component and an Italian component in the Ashkenazi population.

The minority view is that eastern European Jews stem mainly from the Khazars a Turkic people who migrated to the Black Sea area. The debate is over whether the Khazar contribution is major or minor. Elhaik analyzed already collected autosomal data and found a pattern indicating that the Khazar contribution was major.

Most of the posts on this thread lean to the German migration explanation. In addition to my previous comments supporting this view I have located Ellhaik's paper and have noted questionable assumptions of his research. Elhaik assumed that modern Caucasian populations could serve as a proxy for the Khazars of 1000 years ago. On the basis of limited analyses he asserted that the Caucasian populations were genetically differerent from the Semitic populations of the Middle East. This assertion is completely contradicted by Yunusbayev's finding that the Caucasian populations are primarily of Middle Eastern origin. Elhaik also presumed that modern day Palestinians could serve as a proxy for ancient Israelites. While there is significant overlap between Jewish and Palestinian lines, Elhaik seems to be unaware of the Moslem Arabic migration into Palestine long after the Jews had left---e.g. as reflected in Y dna differences in J1/J2 ratios between Jews and Palestinians. On this issue it is not surprising that Jews are more like Caucasians than they are like Palestinians. Lebanese are also more like Caucasians and they certainly are Semitic.

Thanks Josh!

Am I correct in thinking that whether this paper is correct or not that this only explains the origin of Eastern European Jews and not German Jews per-se? If so, do we know the origin of German Jews?

[josh w.]

The choice of Armenians as a proxy for Khazars was particularly dumbfounding.

First, Armenians are from the southern Caucasus, not too far from the northern Near East. A better choice might have been Tatars, Adyghei, Russians from the Volga area, or a similar population.

Second, some diaspora Armenians often have measurable descent from Near Eastern Christian populations who themselves share common ancient ancestry with Jews (e.g. Maronites, Assyrians, Copts). So depending on the exact nature of the Armenian sample, those Ashkenazim might be genetically similar to Armenians because those Aremnians have some Near Eastern influence, rather than the Ashkenazim necessarily having Caucasus influence.

Yes, the comparison should have been between Caucasians and modern day northern Semites such as current Assyrian groups. Data presented at the now defunct Dna Forum demonstrated that Assyrians have Y dna haplogroup distributions that are quite similar to Armenian distributions. My own Y dna J2a* Ashkenazi cluster is shared by a few Maronite Christians---it probably should be renamed the 'Semitic: cluster. I think Lebanese should have served as proxy for the ancient Israelites since the cultures were quite similar and both nations came from Levantine (Canaanite) rather than Arabic stock- there was some Arabic migration to Lebanon but unlike Palestine there is no shared border with Arabia.