So, using nMonte and a database of ancient and modern samples, I tried many models of ancestry for Germany CWC people. Unlike what I have previously read as a hypothesis, a mix of basically Balkans_Chalcolithic + Ukraine_N or a mix of Balkans_Chalcolithic + Ukraine_N + Ukraine_GAC both don't work at all (too high distance rate).
Importantly, all the models (no exceptions) with a much lower (<2%) distance need to have Yamnaya to work (not any other earlier or later steppe sample, far less WHG-like or EHG-like sources to its north). Much more interesting is the fact that the best models all involve Ukraine_GAC as the main source of EEF ancestry in CWC, not EEF of Poland (GAC or TRB), Germany (Germany_MN or Alberstedt_LN) or Sweden (TRB). At least for German CWC it's interesting that TRB (from Poland and from Sweden) always yields 0% if you include Ukraine_GAC. If you include it, too, Poland_GAC yields 0% too.
The best combination of ancestral sources (using several Yamnaya sources as possibilities) I could model and test (distance% = 1.2744) was (for German CWC): 63.5% Yamnaya_Kalmykia, 23.7% Ukraine_GAC, 10.8% Yamnaya_Bulgaria, Narva_Lithuania 2.0%
.Curiously, I tried the same model, using several population references, for Bell Beaker France with lots of steppe ancestry, and its results (also reasonably close - ditance% = 1.1063) are very different from those of German CWC. Would that really be likely if, as Eurogenes thinks, Central European BB were basically a branch of CWC + extra EEF? The results were: Poland_GAC 41%, Yamnaya_Kalmykia 27.5%, Yamnaya_Karagash 15%, Yamnaya_Ukraine 15%, Narva_Lithuania 1.3%, Ukraine_GAC 0.2%. Notice: Poland_GAC instead of Ukraine_GAC as the closest proxy of EEF; and a combination of Yamnaya sources (not including Yamnaya_Bulgaria) instead of almost exclusively Kalmykia+Bulgaria in German CWC.
What do you think those resuls may indicate?
If I'm not making some huge mistake (maybe overfitting?), then it seems like the CWC got almost all their EEF along the way even before mixing with GAC and TRB groups (if they eventually did, I don't know). And it also seems to have got just a little bit more WHG-shifted than what would derive from a simple Yamnaya + GAC mix, possibly owing to the route it took to spread westward in North Europe.