reply to forgetful

Again, thanks for the comment. For info in case didn't come up, there's some more discussion on Eurogenes, in the last 3x threads, and some more observations from myself and others. I've put some in the thread about Khvalynsk and as an off-topic comment on the tail end of "How the Shirenzigou nomads became Proto-Tocharians".  I expanded my attempts at labelling the PCA a bit more: https://imgur.com/a/4coC36k (some of this may be thrown into doubt by your further comment but I think most is consistent).

There are a couple more videos mentioned in that thread (uploaded to the same youtube channel) by Anthony responding to questions from Kristian Kristiansen and others. One in particular talks about the Steppe_Maykop, and how they arise from what Anthony sees as a big expansion in mobility linked to the Mesopotamian metal trade, before Yamnaya, and around the late 4th Millennium. (I think the Glavanesti samples might be particularly interesting in this context; a metals trade link might help explain a fast train movement of Steppe_Maykop like people into Carpatho-Balkan sphere of Romania?). While the other covers why Anthony thinks CHG ancestry enters the Steppe from the Volga Delta.

Good reasoning from comparing the two PCA about which of the Grey Stars are which samples (I'd guess the big cluster probably is more likely to be Romania pre-Yamnaya but it's still not totally clear to me). Yamnaya Ukraine Ozera with her Maykop like ancestry doesn't seem to be on this plot. Anthony talked about samples from Ukraine who have Maykop ancestry at the tail end of his comment on steppe maykop, although whether he meant the mountain flank or steppe variant or both was not very clear.

Yes curious about the Corded Ware Early overlap with Yamnaya Don and early Sredni Stog as well.

Interesting note about the ROU_BA Glavanesti samples and the Yamnaya trail into Carpatho-Balkans including the different reasons why the Yamnaya/"pre-Yamnaya" samples might have those positions which I hadn't considered that.

Looked at it like you suggest, it will need more than 2x dimensions to understand at, to properly distinguish the shifts West:East from being due to CHG or West Siberian or EHG like factors.

Particularly interesting to me is that Anthony notes a direct family relationship between a Steppe_Maykop sample and an Usatovo sample somewhere on the Black Sea. There's already a relationship noted in the latest Human Origins anno file between a Steppe_Maykop sample (SA6010) and one of the "Yamnaya_Caucasus" samples (SA6001), who was then redated to Eneolithic, but showed a mostly typical Eneolithic_Steppe ancestry set.

So the Steppe_Maykop, with this West Siberian ancestry, are perhaps interlinked in some ways to the Steppe societies (Usatovo, Eneolithic people in the steppe), perhaps not a very separate long term population? Probably going a bit far with speculation but it makes me think a bit about how high mobility in migration ages might lead initially to very heterogenous societies, but sometimes this can then be supplanted by a big expansion of a more homogenous and initially small and locally rooted population (heterogenous Migration Period succeeded by "Slavic Expansion", or how ocean trade conducted by sailors from all round the world then leads to settler colonization efforts that are a bit more homogenous in some ways, although that's more of a reach).

Again to understand why these Usatovo samples have the position they do, seems like we'll need more than 2x dimensions. (For example, is the one that's on the Volga cline really on the Volga cline, which is simplest, or is this more like Usatovo+West Siberian ancestry?)

On Anthony's comments about CHG ancestry coming in from Volga-Delta, to me this seems pretty curious if these people were just hunter-fishers and their ancestry was coming at the exact same time as domesticates and copper are being introduced in large quantities. It seems an unusual coincidence, although perhaps not.

....

On another separate note (now for something completely different), in case you're interested (I'd normally put this in an Open Thread):

Couple papers yesterday, showing new line of discussion in the ancient dna genetic stature estimation papers.

1) https://www.biorxiv.org/content/10.1101/2021.03.31.437877v1 - "Predicting skeletal stature using ancient DNA"
2) https://www.biorxiv.org/content/10.1101/2021.03.31.437881v1 - "An integrative skeletal and paleogenomic analysis of prehistoric stature variation suggests relatively reduced health for early European farmers" - suggests "Early farmers were relatively shorter than expected given their polygenic height scores"

Useful comparison because they go beyond the paper by Cox and Iain Mathieson in 2019, which matched general populations polygenic score and recorded height (https://www.pnas.org/content/116/43/21484), and actually match individuals (because the sample sets are so big now that this is feasible to some reasonable extent!). Like that paper, new ones use the stature PRS computed from UK Biobank, to remove confounding of European population structure in GIANT.

These have about the same sample size but seem to use use different sets (paper 2 is based on European samples only).

Doesn't seem based on the idea that scores are perfectly predictable (first paper makes it clear this is not the case) but that "All things being equal".

I thought I'd check them out against population structure on G25, but a lot of the IDs are not the same between the papers and David's set, so it's a bit tiresome.