Actually, the Yemenite Jews are shown to not have SSA admixture in the samples from the study on Post-Roman Egypt. Nevertheless, the regular Yemenite samples are marginally, to considerably intermixed with SSA.
https://www.nature.com/articles/ncomms15694
The Schuenemann paper is really a treasure trove of information, isn't it? Looking at the Admixture graph I'm reminded of why I sometimes have difficulty distinguishing between these groups in terms of phenotype, although I'm sure it's much easier for people from the region.
In many cases the major difference genetically is the intrusion of Yoruba like genes, and the secondary one the amount of Iran Neo/Chl like ancestry. Palestinians and Jordanians are virtually indistinguishable genetically, with Lebanese different in having lower to no Yoruba and slightly more Iran like ancestry. The Druse and the Lebanese look exceedingly similar except for the tiny amount of Yoruba like ancestry. The Algerians and Tunisians have by far the most SSA, followed by Egypt, and have by far the least Iran Neo/Chl. I don't see a heck of a lot of difference between Bedouin A and Egyptians, whereas Bedouin B are very different, with extremely high levels of Natufian, very low levels of Anatolia Neolithic, and no Yoruba.
Looking at the Tunisians and Algerians, I have no idea why anyone would try to use them to model ancient admixture.
Once again we also see how religion can serve as a barrier to gene flow. I would speculate that the Iraqi Jews have some intrusion from Iran Neo/Chl, however. It will certainly be interesting to see first millennium BC and Hellenistic Jews on this admixture graph. I'll really be interested to see how they compare to Lebanese, Palestinians, and the Druse, whom I used to use as a sort of proxy for them.
@Ironside,
The difference in terms of "SSA" between this and gedmatch is probably that on some of those calculator they might be including East African samples. I would think it's probably a better idea to have both reference populations, but then you have to do some math to really figure it out, as I think about 40-50% of "East African" is SW Asian like, which brings down that "SSA" number a lot.
I think you've hit on something important in talking about "dark skin". People assume the darker skin in some Bedouin groups, for example, or Yemeni or Saudi groups is an important marker for how much "SSA" admixture there's been, when that really isn't the case. It's selection once again. Given how much West Eurasian is in Horners, for example, you'd think they'd be lighter skinned, for example.
I think I've asked you this once before, but I unfortunately don't remember the answer. Is their specific data which reliably tells us where the Bedouin A versus Bedouin B samples were taken?