Thread by @veebeeaisdee

Updated PCA of West, East, and North African populations 🧵
Samples below

I'll be commenting on 3 main things related to how we can use this to better understand Pastoral Neolithic populations of East Africa (Cushites)
- Is Natufian truly a good proxy?
- The African cline of ancestry in Cushitic populations
- The "Nilo-Hamitic" cline in Bantu speakers

Looking at the PCA closely, we notice the slope of the "Horn of Africa" grouping ends more southern than Natufians do. We know that this is due to Iberomaurusian admixture, but the question is, do we actually have any Natufian DNA?

Cushitic & Semitic speaking populations of the Horn of Africa have a very consistent percentage of Iberomaurusian ancestry (10.7%, ±1.5%), while Levantine ancestry varies a lot due to Southern Arabian admixture.

There is also a Basal amount of PPNB & Non-PPNB Levantine ancestry, the ratio of Iberomaurusian to Non-PPNB Levantine ancestry being roughly 1-3, while Natufian has a ratio of 1-4. The ratio of PPNB and that mysterious component would be 3-10, which is much higher than expected.

I have also observed this Iberomaurusian to Non-PPNB Levantine ratio of 1-3 in Coptic Egyptian populations which tells us it is probably a real ancient component. which was Diluted by Levantine Farmers as well as Nilo-Saharan pastoralists & East African Hunter Gatherers.

Clearly, this ghost population is real, well, what can it be then? One proposed location where such a population is to be found could be Helwan, an Egyptian city home to a Levantine-associated Late Epipalaolithic culture. We won't truly know until we get more DNA from Egypt.

Now you may be wondering, where does this NiloSaharan sample come from? It was made by extracting the East African ancestry of Mota, and I myself do not understand exactly how it was done, but the sample is so perfect it's almost a miracle.

From this sample we can deduce populations from the Horn of Africa hadn't significantly mixed with actual Nilotes and East African Hunter Gatherers yet, while they started doing so around the SPN areas in Kenya, Tanzania, etc.

We already knew from a while ago that Nilotic populations had a West African like component that was mostly absent in Cushitic and Semitic speaking Horn African populations, but this sample helps us visualize it better.

With modern South Cushites (descendents of SPN samples shown above), we can see there is some sort of Nilo-Hamitic to Bantu cline forming. One Bantu sample, Kikuyu, also makes the cline appear in the Bantu side of things. We also notice a Proto-Nilotic to West African cline

Riverine populations from Sudan are more southern shifted than Cushitic-Semitic Horn African populations, likely due to admixture with unknown Saharan and Nile valley populations, which the Beja mostly lack. (G25 run not fully accurate, but it illustrates what I'm showing)

There are still a few things that I am confused about. For example, what led to the dissapearance of this purer Nilo Saharan population (check empty triangle), and what exactly makes up Nubian ancestry. But I hope this thread cleared up some things about East African Genetics.

Samples used:
pastebin.com/JJASDbFR

@threadreaderapp unroll