Substructure in EEFs

[JRD]

There's some substructure in EEFs. There are two main clades of EEFs, which are Iberian EEFs and Danubian EEFs. The main difference between them are the presence or absence of Neolithic Iranian ancestry (which seems closest to IRN_Wezmeh_N), yet there are some EEFs in Central and Eastern Europe lacking that ancestry. Chief among the latter group is UKR_N_o:I3719, especially since it's the easternmost EEF sample known (at least of its time) and representative of an EEF population that contributed ancestry to Steppe groups. Then, there's variation in the amount of what looks like CHG ancestry, along with variable amounts of Natufian ancestry. Plus, there's local HG admixture, whose amount and nature varies on location and time period. Notably, variations in Neolithic Iranian, CHG, and Natufian admixtures are already seen in Neolithic Anatolia, with TUR_Barcin_N:I1096 as the only known Ceramic Neolithic-period ANF lacking Neolithic Iranian ancestry. The substructures in the founding ANF populations and their EEF descendants could explain the differences between Iberian and Danubian EEFs (besides different HG admixtures), the extra Neolithic Iranian ancestry in Italian EEFs, and the lack of it in some Central and Eastern European EEFs.

[JRD]

To continue with the above, TUR_Kumtepe_N has the highest amount of Neolithic Iranian ancestry, followed by GRC_Peloponnense_N, in turn followed by TUR_Tepecik_Ciftlik_N, then by ITA_Ripabianca_Di_Monterado_N. Both Italian EEF groups are variable in the amount of Neolithic Iranian ancestry, though all of the ITA_Ripabianca_Di_Monterado group have more of it than TUR_Barcin_N, while only two of the ITA_Grotta_Continenza_N group have more of it than TUR_Barcin_N. Using TUR_Barcin_N as a source produces different results. The Central and Eastern European samples lacking Neolithic Iranian ancestry are (in alphabetical order) DEU_LBK_SMH:XN173, HRV_Sopot_MN:I5077, HRV_Impressa_N:I5071, HRV_Impressa_N:I5072, HUN_Lengyel_LN:I1899, HUN_Vinca_MN:I1887, HUN_ALPc_Szakalhat_MN:I2745, HUN_Protoboleraz_LCA:I2789, and UKR_N_o:I3719. Interesting about that set of individuals is that they have almost the full array of Y-DNA haplogroups of Neolithic Anatolia (J2a being the missing haplogroup), and that TUR_Barcin_N:I1096 (which belongs in that group despite being in Anatolia) is the I2c sample of the site. Most also show autosomal evidence of HG admixture, and there are HG haplogroups in two of them, namely DEU_LBK_SMH:XN173 (who has mtDNA U5a2d) and UKR_N_o:I3719 (who has Y-DNA I2a2a). The presence of CHG ancestry independently from Neolithic Iranian ancestry points to two separate dispersals (the former possibly from Northern Anatolia, the latter from Eastern Anatolia). This was all done by Vahaduo models (no screenshots available) using TUR_Pinarbasi_HG, IRN_Wezmeh_N, Levant_Natufian, GEO_CHG, and local HGs as sources, then using more proximal sources, and drawing conclusions from the results of both distal and proximal source models for most of the EEFs. The haplogroup data of some of the EEFs were searched differently.