27 January 2009

DNA: New study predicts Ashkenazi Jewish ancestry

A genetic signature of Jewish ancestry perfectly separates individuals with and without full Jewish ancestry in a large random sample of European Americans, according to a new study in Genome Biology (January 2009).

Authors are Anna C Need, Dalia Kasperaviciute, Elizabeth T Cirulli and David B Goldstein. Goldstein authored Jacob's Legacy (click here).

The researchers conclude:
within Americans of European ancestry there is a perfect genetic corollary of Jewish ancestry which, in principle, would permit near perfect genetic inference of Ashkenazi Jewish ancestry. In fact, even subjects with a single Jewish grandparent can be statistically distinguished from those without Jewish ancestry.
The team further concludes that in the context of the sample studied:
... 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.
Download the study here, with charts available as separate files. This is a slow download - be patient.

Why this study? The paper indicates that although it was recently shown that genetic distinction between self-identified Ashkenazi Jewish and non-Jewish individuals can be seen in European American genetic patterns, no study had shown how accurate was that self-identified Ashkenazi Jewish ancestry, or the degree of Jewish ancestry among individuals with fewer than four Jewish grandparents.

The random sample included 611 unrelated self-described Caucasian subjects - most in America - who specifically reported whether they had Jewish ancestry, and if so, how many grandparents were Jewish. Each participant was genotyped for some 550,000 polymorphic markers. See the study for the breakdown of the group which included a very tiny minority of Sephardim. Most Jews were Ashkenazi.

The study indicated that every participant with self-reported full Jewish ancestry had a higher score than an individual with no Jewish ancestry. Sephardic participants were only only a few, so results cannot provide information on Sephardic or Mizrachi Jews.

In one paragraph, it appears researchers are not aware of Jewish history or migration - I was surprised that Goldstein did not further explain this quote in the study.
The majority of informative subjects with no Jewish ancestry that scored most highly on PC1 were either of Italian or Eastern Mediterranean descent.
To me, it is obvious why these particular participants with "no Jewish ancestry" scored so high. Following the expulsions from Spain (1492), Sicily (1493) and later Portugal, the Jewish population of Sicily was the largest in the world. Following that expulsion, Jews crossed the Straits of Messina into Calabria where the population was estimated as 50% Jewish. I also believe that the term Eastern Mediterranean actually refers to Greece (more specifically Rhodes and Salonika) and Turkey (Istanbul and other cities), which were major destinations for Sephardic Jews following these expulsions.

According to the article, the analysis makes clear that those with full Jewish ancestry are genetically distinct from those having no self-reported Jewish ancestry. Those who self-identified as Jewish and knew their origins were nearly all Ashkenazim. Of the Jewish participants who didn't know whether they were Ashkenazim or Sephardim, but could say where a grandparent had come from, most had Eastern or Central European roots. None had Mediterranean or Middle Eastern ancestry.

Read the complete study at the link above.


  1. For anyone with a knowledge of statistics and modeling as used in genetics, I would be interested in hearing what you think of any prediction instrument that has 100% sensitivity and 100% specificity. As one with a statistical background (but not in genetics), I have never seen a screening test or diagnostic test with 100% sensitivity and specificity (or, as the correct terms here would be, predictive value positive and predictive value negative).

  2. I am not a geneticist nor a statistician, but it seems strange for a group of scientists to claim 100% sensitivity and 100% specificity. Other writers and readers have also addressed this point. If this is true, then I'd love to see a similar test for Sephardic ancestry!

  3. Anonymous1:16 PM

    There are other possible reasons for genetic similarities between Jews and a proportion of East Mediterraneans and Italians. One would be Middle Eastern population movement into Europe during neolithic times . The theory was proposed by Prof. Luigi Luca Cavalli-Sforza at Stanford. Although he overestimated the neolithic genetic component in Europe overall, there is a clear clinal signature of neolithic expansion into Europe that is highest in the southeast (Greece, Balkans) and declines towards the northwest. The second would be intermarriage between Jews and local populations during the period of the Roman Empire. There were sizable Jewish communities in Roman Alexandria, western Anatolia, Greece, and Italy.

  4. Anonymous7:33 PM

    Abraham's Children is a journey back in time and through history to reconnect with our ancestors.
    The Jewish Journal of Los Angeles reflects on Abraham's Children's unique take on Jewish identity...

    Hillel Halkin writes that "Entine has done a serious and responsible job of reporting" on "Jews and their DNA" in the September Commentary magazine

    Read Jon's cover story in Ancestry Magazine about the problematic consequences of DNA research into human differences

    Library Journal: "[Jon Entine's] explorations take him from his own Jewish family members' cancer diagnoses to genetic labs—both academic and commercial—across the world, with lengthy forays into Jewish history, the history of Israel and Zionism, and the split between social anthropology and biological anthropology."


  5. I'm always happy to point people to Jon's articles and appearances. I first read Abraham's Children in galley proof when I met Jon at an international Jewish genealogy conference. His views are always interesting and enlightening.

  6. Aaron Merrari7:58 PM

    The study reviewed is typically and incredibly Ashenazocentric. They write "European" when they mean mostly Eastern and part of Central Europe-pretty arrogant considering that Greek and Roman Jewry predated there being any such thing as an Ashkenazim by 12-15 hundred years, and Rome has the longest continuous community in all of Europe! Wake the f#@% up Jew=Ashkenazi is as racist as the pre-Civil Rights white gentile attitude toward blacks. What year are these people living in?!

    In any event, the last study I saw showed only 50% Jewish ancestry in most edoth* studied-except the ones that were lower, so 100% is a hubristic and ludicrous number We know little if anything solid about the 9 1/2 tribes of the Northern Kingdom, and if you want to go by the Rabbinic decision to switch to matrilineal descent**, Dinah's descendants.
    *Socio-geographic communities Yemenite, Iberian, Roman, Ethiopian, Eastern European, etc.
    **I believe that decision was dubious-possibly overstepping the mandate for the "chachamim" to change what was indicated by the written Torah. Considering that Genghis Khan has at least 17 million patrilineal descendants, the implications I find staggering. Maybe a billion Jews by real p'shat Torah reckoning? At least a half-billion. I think one has only to look at those they know (and even more so Ashkenazim)to see that the fathers genes tend to be dominant more often than not-in either sex of offspring. Peace

  7. Anonymous7:04 AM

    It really seems very odd that one DNA-study says something else. And I am really convinced that we will never know the the whole facts. But it is clear when one reads most of the respected DNA-studies. That the Asheknazi have and admixture of European and Middle Eastern DNA. But the question remains witch one is the most?