When trying to learn and understand approximately everything, one is forced to periodically admit that there are a great many things one does not yet know.
I made a diagram of my thoughts from yesterday:
Abbreviations: SSA = Sub-Saharan Africa; ANE = Ancient North Eurasian, even though they’re found all over the place; WHG = European hunter-gatherers; I-Es = Indo-Europeans.
I tried my best to make it neat and clear, focusing on the big separations and leaving out the frequent cross-mixing. Where several groups had similar DNA, I used one group to represent the group (eg, Yoruba,) and left out groups whose histories were just too complicated to express clearly at this size. A big chunk of the Middle East/middle of Eurasia is a mixing zone where lots of groups seem to have merged. (Likewise, I obviously left out groups that weren’t in Haak’s dataset, like Polynesians.)
I tried to arrange the groups sensibly, so that ones that are geographically near each other and/or have intermixed are near each other on the graph, but this didn’t always work out–eg, the Inuit share some DNA with other Native American groups, but ended up sandwiched between India and Siberia.
Things get complicated around the emergence of the Indo-Europeans (I-Es), who emerged from the combination of a known population (WHG) and an unknown population that I’m super-speculating might have come from India, after which some of the I-Es might have returned to India. But then there is the mystery of why the color on the graph changes from light green to teal–did another group related to the original IEs emerge, or is this just change over time?
The IEs are also, IMO, at the wrong spot in time (so are the Pygmies.) Maybe this is just a really bad proxy for time? Maybe getting conquered makes groups combine in ways that look like they differentiated at times other than when they did?
Either way, I am, well, frustrated.
EDIT: Oh, I just realized something I did wrong.
Among other things, I realized I’d messed up counting off where some of the groups split, so while I fixing that, I went ahead and switched the Siberians and Melanesians so I could get the Inuit near the other Americans.
I also realized that I was trying to smush together the emergence of the WHG and the Yamnaya, even though those events happened at different times. The new version shows the WHG and Yamnaya (proto-Indo-Europeans) at two very different times.
Third, I have fixed it so that the ANE don’t feed directly into modern Europeans. The downside of the current model is that it makes it look like the ANE disappeaed, when really they just dispersed into so many groups which mixed in turn with other groups that they ceased existing in “pure” form, though the Bedouins, I suspect, come closest.
The “light green” and “teal” colors on Haak’s graph are still problematic–light green doesn’t exist in “pure” form anywhere on the graph, but it appears to be highest in India. My interpretation is that the light green derived early on from an ANE population somewhere around India (though Iran, Pakistan, the Caucuses, or the Steppes are also possibilities,) and somewhat later mixed with an “East” population in India. A bit of that light green population also made it into the Onge, and later, I think a branch of it combined with the WHG to create the Yamnaya. (Who, in turn, conquered some ANE groups, creating the modern Europeans.)
I should also note that I might have the Khoi and San groups backwards, because I’m not all that familiar with them.
I could edit this post and just eliminate my embarrassing mistakes, but I think I’ll let them stay in order to show the importance of paying attention to the nagging sense of being wrong. It turns out I was! I might still be wrong, but hopefully I’m less wrong.