Fails to adress the likely Anatolian origin Luis Aldamiz, none 16 December 2009 I must say that I'm again disappointed at the approach of the paper. Arabs and Europeans are used as controls but a large void is left (again) in the area around modern Turkey, where the ancient Jewish Diaspora lived for many centuries (in a time when proselytism was common in Judaism too). I find this most striking but also systematic failure of Jewish genetic studies in general. Only studies no focused in Jewish ancestry, like Bauchet 2007, have provided some clarification on this aspect, clustering Jews with Greeks and Armenians. But they await to be confirmed. The first thing I looked for in this paper was whether Turks (or maybe Balcanic peoples) had been sampled as controls. Sadly it was not the case. Otherwise the paper does provide some interesting information like the clustering of diverse Jewish groups and their lack of clustering with any of the other groups, notably Palestinians and Druzes, in their main component. But I really believe that using better chosen control populations, Turks among them, would give much more clear and interesting results. Competing interests None in particular.