Trying to find clues to the origin of "Forager A", the non-Fertile Crescent element in Europeans and Caucasians, I decided to run an unsupervised analysis of all Northern Eurasian populations. I don't have much faith in components from unsupervised mode actually corresponding to closely related ancestral populations just because ADMIXTURE decided on it's own it's so. Still unsupervised runs are great to get a picture of relationships.
I also included the San, since I wanted to see if they were actually related to "Forager A", Hadza, Mbuti Pygmy, !kung, and Papuans, Melanesians. And why not, also the East Asian populations because of the proposed Altai language family and it's tenuously possible links to Korean and Japanese. And since I was at it, I also threw in those often forgotten populations of the "Siberian" belt, Amerindians.
Remember this is an unsupervised run. Some noise <1% is expected, but I think >1% consistent components across similar populations are significant. I won't immediately name the components ADMIXTURE came up with. Also, these are kind of chunky, particularly for smaller components, so this is just a simple experiment I wanted to share, to get hints about what "Forager A" might actually be, and also for my future East Asian Neolithic run.
"Forager A" is about Amerindian women or slaves brought to Europe a few centuries ago by the way... At least it's hard to explain how they would get to Dagestan.
I think "Forager A" is likely "Siberian"-like. As for closer affinities among "Siberians" it could be drift/founding effects making more distantly related populations seem closer? Seems unlikely and there are other theories considered unlikely as well...
"Yoruba" in Druze and others is I think attracting a bit of the Nile Core Component (Northeast African) by the way, since these could be distantly related via the Green Sahara.
This is an unsupervised run though. It is quite chunky. These are small elements.
For now, I'll just say that "Forager A", if thats what unsupervised ADMIXTURE is picking up, looks interesting indeed.
I'll also be doing a supervised, hopefully more homogeneous run.
I also ran East Asian populations in the same run, I'll present them in the next post tomorrow.