I know Copper Age Italy didn't have steppe, thus the qualifier that R437 does have some steppe ancestry and it's not exactly a two-way model. Also, here's the table from the supplementary section showing their two-way model:
And from the published paper:
Now, while I don't completely trust G25, I made a few models based on R437 anyway to compare:
The Antonio et al model:
Target: ITA_Prenestini_tribe_IA_o
Distance: 3.3803% / 0.03380304
50.4 ITA_Grotta_Continenza_CA
49.6 ARM_LBA
Here I add a steppe component and switch out the chalcolithic sample with a neolithic one (it admittedly didn't work with the CA sample as it rejected the steppe component):
Target: ITA_Prenestini_tribe_IA_o
Distance: 2.5148% / 0.02514792
57.4 ITA_Grotta_Continenza_N
32.4 ARM_LBA
10.2 Yamnaya_RUS_Samara
Using IA Anatolia the steppe component doesn't get rejected with CA Italy but offers a worse fit:
Target: ITA_Prenestini_tribe_IA_o
Distance: 4.7940% / 0.04794002
54.0 ITA_Grotta_Continenza_CA
36.6 TUR_IA
9.4 Yamnaya_RUS_Samara
Even though the paper didn't mention Bronze Age Anatolia in their model, I tried it anyway since you mentioned it. Curiously it becomes to dominant element, although I'm not sure how true that is in reality:
Target: ITA_Prenestini_tribe_IA_o
Distance: 1.9887% / 0.01988730
53.8 TUR_Isparta_EBA
30.8 ITA_Grotta_Continenza_CA
15.4 Yamnaya_RUS_Samara
Target: ITA_Prenestini_tribe_IA_o
Distance: 2.2060% / 0.02206050
51.2 TUR_Kaman-Kalehoyuk_MLBA
33.2 ITA_Grotta_Continenza_CA
15.6 Yamnaya_RUS_Samara
The question remains right now when and how this Anatolian/Caucasus ancestry came into this part of Italy (and in the south as well). Hopefully we'll get ancient samples from mainland southern Italy sooner rather than later that can shed more light on this.
Bookmarks