Gene Expression

How Ashkenazi Jewish are you?

Carl Zimmer pointed me to a new paper, A genome-wide genetic signature of Jewish ancestry perfectly separates individuals with and without full Jewish ancestry in a large random sample of European Americans. The title is so informative that pasting the abstract is almost unnecessary, but here is the conclusion which gets to the point:

In conclusion, we show that, at least in the context of the studied sample, it is possible to predict full Ashkenazi Jewish ancestry with 100% sensitivity and 100% specificity, although it should be noted that the exact dividing line between a Jewish and non-Jewish cluster will vary across sample sets which in practice would reduce the accuracy of the prediction. While the full historical demographic explanations for this distinction remain to be resolved, it is clear that the genomes of individuals with full Ashkenazi Jewish ancestry carry an unambiguous signature of their Jewish heritage, and this seems more likely to be due to their specific Middle Eastern ancestry than to inbreeding.

There have been other papers which show that Ashkenazi Jews form a separate cluster from gentile whites in the United States. This is important again in the context of biomedical studies attempting to ascertain the genetic roots of particular diseases; population substructure (e.g., Jew vs. non-Jew) may result in confounded associations. Also, one of the authors of the paper is David Goldstein, author of the fascinating Jacob’s Legacy.

In any case, on to the PC charts where the real action is. Do note that I’ve resized and added explanatory labels here & there for clarity.

i-a2f589da83c5fdb38236e67105526c0c-jewpc1.jpg

As you can see, there is almost perfect separation between the Jewish and non-Jewish clusters here. PC 1 = first principle component of variation, and PC 2 second principle component of variation. These are the two largest independent dimensions of genetic variance extractable out of the data set. The authors note that there is almost perfect separation along PC 1, and, they note that most of the gentile whites who are closest to the Jews on this PC are of Italian or Eastern Mediterranean origin. This is important later.

i-514e27460f3beec76649da58e082aaf8-jewpc2.jpg

As you can see here, the interesting point is that Jewish ancestral quanta is roughly predictive of genetic position. This shouldn’t be that surprising once we know that Jews and non-Jews separate so cleanly (e.g., someone who is biracial would be located between their two parent racial clusters on any plot), but it is striking nonetheless in reaffirming the genetic reality of Jewishness. The United States has an easy genealogical history for Jews, as the ancestors of all self-identified American Jews today arrived on the order of 4-6 generations ago. These individuals were likely Jewish relatively far back in their lineages, and admixture is easy to recall because it is of recent vintage.

Finally, let’s put Ashkenazi Jews in the worldwide context.

i-a0f9bdcea9f789e8de3f266c708a7d7d-jewpc3.jpg

I labeled a few populations for clarity. Orcadians are individuals from the Orkney Islands, just off the northern coast of Scotland. As you can see, American whites occupy an expansive region, in large part due to their diverse origins and admixtures. The Druze have been genetically isolated for over 5 centuries, so it is not particularly surprising they form a distinct cluster. It is likely a function of drift due to bottlenecking, just like Iceland. The separation between the Ashkenazi and the Palestinians might be of interest to some because of political reasons.

As the authors note, the Ashkenazi are almost certainly a compound population. There are full-Jews who “look gentile” and full-Jews who could pass easily in the Middle East as a native. Most Ashkenazi Jews exhibit a mix of features. The genetic likelihood that the Ashkenazi have varied origins should no surprise. The paper notes that Askhenazi heterozygosity is actually somewhat greater than American whites, implying that the source of their distinctiveness has to do with genetic origin, not genetic history (i.e., population bottlenecks). Another population which I suspect resembles the population history of the Ashkenazi are Uyghurs, who fall between Europeans and Han Chinese in their ancestry. But the hybridization even occurred in the past, so the Uyghurs today can be thought of as a separate population with its own suite of genetic combinations, as opposed to a hybrid of two “pure” groups. Similarly, Ashkenazi Jews no doubt emerged from hybridization evens between Middle Eastern and European females, but over the generations they have been relatively endogamous (most gene flow has been out of the community into gentile Europeans through conversion) and so are a genetically coherent population in their own right.

Note that the authors had only a few non-Ashkenazi Jews in their sample, so the assertions above only apply to the Ashkenazi (95-99% of America’s Jewish population, but only ~50% of Israel’s).

Update: Dienekes emphasizes that the hypothesis that the admixture element which makes Ashkenazi Jews distinctive from other Europeans is Middle Eastern is somewhat speculative. I’m less skeptical than him seeing as how some Y lineages are Middle Eastern, but he offers this interesting point:

With that said, I do suspect that the distinctiveness of the Ashkenazi Jews is in part due to the possession of a Middle Eastern component of unspecified strength. I base this hypothesis on the results reported to me about the EURO-DNA-CALC test. This test distinguishes between NW, SE Europeans and Ashkenazi Jews; a few Arab individuals who have communicated their results to me have reported fairly high AJ components, indicating that part of what distinguishes an AJ from Europeans is related to the Middle Eastern Semitic background of that group.

Comments

  1. #1 Charles Iliya Krempeaux
    January 22, 2009

    What about a person who has no Ashkenazi ancestry, but has mixed European and Middle Eastern ancestry? Would they seem Ashkenazi to this test?

    (Or does the test look from some allele(s) that are somewhat unique to Ashkenazi, and not European or Middle Eastern populations in general.)

  2. #2 HadasS
    January 22, 2009

    My first thought was “Great, next time they will be able to kill us more accurately.”

    Yeah, I’m paranoid, why do you ask?

  3. #3 razib
    January 22, 2009

    What about a person who has no Ashkenazi ancestry, but has mixed European and Middle Eastern ancestry? Would they seem Ashkenazi to this test?

    yes. i think so. remember, southeast europeans were closest to the ashkenazi on PC 1 for whites.

    (Or does the test look from some allele(s) that are somewhat unique to Ashkenazi, and not European or Middle Eastern populations in general.)

    no, i don’t think so. papers is 550,000 markers.

    My first thought was “Great, next time they will be able to kill us more accurately.”

    well, fewer false positives ;-) (bright side of things?)

  4. #4 Joshua Zelinsky
    January 22, 2009

    This seems to go against the common sense version of Askhenazic history. As it is normally told, one would expect a lot of bottlenecking. It isn’t obvious to me how to reconcile them. I’m surprised that that standard Ashkenazic story is that wrong. On the other hand, this does confirm the view that Askenazim are really Middle Eastern and not just a collection of Europeans who became Jewish late.

  5. #5 razib
    January 22, 2009

    On the other hand, this does confirm the view that Askenazim are really Middle Eastern and not just a collection of Europeans who became Jewish late.

    well, part middle eastern. i mean, mestizos are part european and part amerindian, but something new. in any case, this isn’t the first or only paper suggesting these conclusions.

  6. #6 Charles Iliya Krempeaux
    January 22, 2009

    @Joshua Zelinsky, you said…

    This seems to go against the common sense version of Askhenazic history. As it is normally told, one would expect a lot of bottlenecking. It isn’t obvious to me how to reconcile them. I’m surprised that that standard Ashkenazic story is that wrong. On the other hand, this does confirm the view that Askenazim are really Middle Eastern and not just a collection of Europeans who became Jewish late.

    Well, we could get into the theories that Ashkenazi are descendant from Khazars :-)

    Would modern day Persians in north west Iran, Azerbaijanis, Georgians, Armenians, Ossetians, etc be good substitutes for Khazars? Maybe we could put them on the PC map too, and see where they get placed.

  7. #7 razib
    January 22, 2009

    the adygei are from the caucasus. but i doubt that means much, probably similarities due to south-north gene flow over time.

  8. #8 ben g
    January 22, 2009

    here’s my summary of (Ashkenazi) Jewish genetics that i give to lay people:

    Ashkenazi Jews are a unique hybrid Middle-Eastern and (Eastern) European group. As a result, they are white, and specifically they group best with Northern mediterraneans like N. Italians, Greeks, etc.

  9. #9 Charles Iliya Krempeaux
    January 22, 2009

    @ben g, you said…

    and specifically they group best with Northern mediterraneans like N. Italians, Greeks, etc

    According to the PC map, they seem to group better with the Adyhei. And thus perhaps might also group with the Persians in north west Iran, Azerbaijanis, Georgians, Armenians, Ossetians, etc (if they were on that PC map too).

  10. #10 henry harpending
    January 22, 2009

    Joshua Zelinsky points out that the paper contradicts the idea that a bottleneck or several bottlenecks contributed to Ashkenazi genetics. There never was a shred of evidence for any bottleneck at all–the idea gained currency as an explanation for the funny disease picture. This paper, along with several others like the Olshen et al. paper, ought finally to rid us that idea.

    My take on the uniqueness of the pure Ashkenazi sample is that admixture with Europeans happened early on, in Roman times or shortly after, and that since then there has been hardly any.

    Interesting that they essentially replicated and found the same thing that Greg and I did in our Ashkenazi paper, only with a thousand times as many loci.

  11. #11 gcochran
    January 22, 2009

    You’re wrong. It’s Middle Eastern plus _Southeastern_ Europeans: closer to Italians than Russians or Poles, for sure. No signs of much Eastern European admixture.

  12. #12 ART
    January 22, 2009

    From the excerpt: “…seems more likely to be due to their specific Middle Eastern ancestry than to inbreeding.”

    Not sure if that is correct. The Ashkenazim are quite obviously a very inbred population – just look at how many genetic diseases/disorders they carry: http://en.wikipedia.org/wiki/Ashkenazi#Specific_diseases_and_disorders

    And that’s only a partial list…

  13. #13 razib
    January 22, 2009

    The Ashkenazim are quite obviously a very inbred population – just look at how many genetic diseases/disorders they carry

    there’s an alternative explanation.

  14. #14 Mencius
    January 23, 2009

    Quadroon Jews! I’m glad to see the subject raised, because my daughter is one – though I usually shorten it to “Jewroon.”

  15. #15 David Boxenhorn
    January 23, 2009

    Joshua,

    This seems to go against the common sense version of Askhenazic history. As it is normally told, one would expect a lot of bottlenecking. It isn’t obvious to me how to reconcile them.

    Here’s how. You have to take into account the small long-term effective population sizes of Europeans. For the most part, Europeans have been living in the same place breeding with their neighbors for a long time. I bet if you calculate Ne over the past 20,000 years for Europeans it would be pretty small.

    In contrast, Ashkenaz is a hybrid of two widely separated populations, so a small population would be relatively heterogeneous.

    Henry,

    Here is the scenario that I’ve never understood how you explain away: Part of the population underwent a bottleneck, then mixed with a population which didn’t undergo a bottleneck.

    If for example, 50% of the current population’s ancestry descends from a bottlenecked population, rare recessives would be half of what they originally were – a high number, but overall heterogeneity in the genome would still be high due to the 50% non-bottlenecked ancestry.

  16. #16 toto
    January 23, 2009

    Strangely enough, in previous studies on the Y chromosome, Jews (including Ashkenazi and Roman Jews) were strongly clustered with non-European Arabs…

    http://bhascience.blogspot.com/2009/01/shared-genetic-heritage-of-jews-and.html

  17. #17 Tom
    January 23, 2009

    How do the claims of a Khazar origin for Ashkenazis fit with the research? Would people from the historic area of Khazaria test as Middle Eastern?

  18. #18 Mark
    January 23, 2009

    I forwarded this to a friend of mine whose mother was adopted by a Jewish family and raised Jewish. There is some speculation in his family that his mother’s biological mother was also Jewish and the adoption was arranged through religious channeles, but no one surviving knows for certain. Also, there’s some speculation that his father’s family converted to Christianity during the 1930s, when all that… unpleasantness… in central Europe was going on, but again, no one alive seems to know definitively.

    Anyway, he wants to know if there’s now a commercial genetic test he could do to see if he has any actual Jewish ancestry?

  19. #19 razib
    January 23, 2009

    Strangely enough, in previous studies on the Y chromosome, Jews (including Ashkenazi and Roman Jews) were strongly clustered with non-European Arabs…
    http://bhascience.blogspot.com/2009/01/shared-genetic-heritage-of-jews-and.html

    the authors above allude to that finding. they assume that hybridization occurred through male jewish + female gentile pairings.

    How do the claims of a Khazar origin for Ashkenazis fit with the research? Would people from the historic area of Khazaria test as Middle Eastern?

    the proper response to the first question would be to look in wikipedia ;-) khazars were turkic.

  20. #20 Huxley
    January 23, 2009

    I always thought that Ashkenazi Jews look like a cross between Armenians and Europeans.

  21. #21 patrick
    January 24, 2009

    What razib said about southeastern Europeans being close to AJ’s on PC1 was interesting. That could explain a Sicilian like me being mistaken for Jewish (more than a few times). LOL.

  22. #22 Hopefully Anonymous
    January 26, 2009

    Surprised this hasn’t been asked yet. Does this give any insight into the IQ differential? It’s not like the two smartest non-ashkenazi populations have been known to be Italians and Arabs (to my knowledge).

  23. #23 dumbirish
    January 27, 2009

    Hopefully Anon- “It’s not like the two smartest non-ashkenazi populations have been known to be Italians and Arabs” You must be half Italian and half Arab.

    I always thought big foreheads denoted intelligence until I saw Pat Robertson and his biggerheaded son on tv; biggest heads,stupidest people on the planet.

    We need to get some of these ashkenazi guys to go down to Wall Street and straighten this stuff out.

  24. #24 ivy
    January 27, 2009

    Not sure if you’ll read this Hopefully Anonymous, but I’ll give it a try. (I’m an engineer so excuse the biological glibness).

    I believe the Middle Easterners who immigrated to Europe were likely no different than other Middle Eastern ethnic groups. Following significant inter-breeding with Southeastern Europeans, the Jews, probably for religious reasons, broke off and engaged in almost exclusive in-breeding. I believe around 200 A.D., Jewish culture turned intellectual. The reason was probably just random but masculinity turned from the dominant Roman paradigm of laborious physical work, to being able to withstand long hours of Talmud study. So it was clearly an almost artificial selection process rather than the natural ones that resulted in the racial disparities we observe today.

    Did you read this? If someone posts on an old thread and no one’s around to read it, does it make a sound?

  25. #25 Ponto
    January 28, 2009

    What I see from the world wide diagram is that the Ashkenazim Jews tack onto to the end of the European caucasoid groups as do the Adygei, and the Jews/Adygei are separated clearly from Middle Easterners. What that says to me is Jews are just a bunch of European caucasoids and not Middle Eastern caucasoids.

    SNP studies now can separate Norwegians from Swedes as they are that fine. I am sure parents and their children will be separated as effectively on SNPs. It is to be expected that a group of people who separate themselves from others based on a shared mythology and religion would be different from other Europeans on SNPs.

    Jews are just a self separating group of Europeans.

  26. #26 Hopefully Anonymous
    January 28, 2009

    Ivy, I’ve read that narrative before. But at this stage it seems to me to be a coin toss whether or not that’s a just-so story. I’m hoping for more expert insight on this blog.

  27. #27 Matt
    January 28, 2009

    This is fairly tangential, but I’m surprised at the relatively higher position in the north west direction of the sample of “American Whites” relative to Europeans, as well as the fact that American Whites cover more or less the whole area of the European plot, given that the European samples were gathered from virtually across the entire continent and include representatives of Southern European, Central European and Northern European populations. I’m assuming this is just a consequence of different sample size, but it kind of makes me go “Huh?” all the same.

  28. #28 Pierce R. Butler
    January 29, 2009

    I’d be most interested in reading Razib’s take on this book in light of the above data.

  29. #29 Dub Samurai
    February 1, 2009

    I always thought that ashkenazi jews result from the mixture of ancient hebrews with russians and germans, and by such possess a more european look, while sephardi and mizrahi jews have preserved better their middle eastern looks.

    I’m not a specialist but i’m guessing 10% of the first ashkenazi getting mixed with germans and russians would be enough to change the physcal aspect of the rest of the ashkenazi population in the centuries to come, even if the intermixing process became marginal or even null untill today.

    On the other hand, when i see iranians and afghans with blonde hair and blue eyes i wonder if some of the jews who travelled to central and eastern Europe were not also mixed with ancient persians? The same kind of doubt occurs to me when i look at some people in North Africa, namely tunisians and moroccans who have blue eyes.

    Finally i’ve read some studies saying most genetic studies indicate that most jews, according to the Y chromossome analysis fall into the J haplogroups, meaning they’re mostly a semitic people. Is this true? What about mtDNA tests? How credible are these tests?

    I would like to see some genetic tests on modern palestinians though. Are they too descendants of ancient hebrews who converted to Islamism? They have a strong arab influence in genetical terms for sure, but i have a strong guess many of them also descend form ancient hebrews, cannanites, jebusites and other ethnic groups who lived in that region before the arab occupation.

  30. #30 Dub Samurai
    February 1, 2009

    I apologize for the grammar errors and for some idiot quesitons? The genetic test are credible. It’s the statistical conclusions that seem doubtfull to me. I mean, how large do samples need to be in order to reach a credible conclusion?

    By the way my question about an eventual mixture of ashkenazi and persians also leads me to wonder if there is no need for jews to be mixed with europeans in order to see some jews with blonde hair and blue eyes. The reason for this lays in the fact that you have some iranians, afghans, pashtuns and even indians with blonde hair and blue eyes looking pretty much caucasians. Or is it that these blonde and blue eyed people are a product of mixture with european groups (maybe russians)? Perhaps they are direct descendants of some of the ancient indo-europeans tribes.

  31. #31 James
    February 19, 2009

    Well I propose a toast to hybrid vigor!

    http://images.allposters.com/images/71/039_35561.jpg

  32. #32 Bob Bollow
    March 7, 2009

    Is there any way to read the data to suggest that the Ashkenazi Jews are not of Middle Eastern descent? I’m wondering if the idea that Ashkenazi are of middle eastern decent is a preferred way to read the information, so is this a theory or certainty?

  33. #33 Frances
    April 10, 2009

    What can anyone tell me about a DNA subject of my family whose results are Haplogroup Q – clearly markers back to Belgium, Rumania, Latvia, etc. Interestingly, he is of Choctaw tribe. There was a Trader Huet/Hewitt documented in MS Choctaw Nation at a specific village as resident trader at the time his ancestor was born c1771. What is his paternal ancestor’s most likely entrance to this country in order to find a young man of Askenazi origins in the Choctaw Nation c1771? By way of LA French colonials early 1700s? By way of Charleston with Palatine Huguenot immigration of approx 75 families by 1700? West Indies?

    Results on 12 markers: 13, 23, 13, 10, 14, 16, 12, 12, 11, 14, 15, 30
    These results were one marker off from a subject of India. Explanation? Too far apart to count?

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.