On the limits of fitting complex models of population history to f-statistics:
These results show that at least with regard to the AG analysis, a key historical conclusion of the study (that the predominant genetic component in the Indus Periphery lineage diverged from the Iranian clade prior to the date of the Ganj Dareh Neolithic group at ca. 10 kya and thus prior to the arrival of West Asian crops and Anatolian genetics in Iran) depends on the parsimony assumption, but the
preference for three admixture events instead of four is hard to justify based on archaeological or other arguments.Why did the Shinde et al. 2019 AG analysis find support for the IP Iranian-related lineage being the first to split, while our findGraphs analysis did not? The Shinde et al. 2019 study sought to carry out a systematic exploration of the AG space in the same spirit as findGraphs—one of only a few papers in the literature where there has been an attempt to do so—and thus this qualitative difference in findings is notable. We hypothesize that the inconsistency reflects the fact that the deeply-diverging WSHG-related ancestry (Narasimhan et al. 2019) present in the IP genetic grouping at a level of ca. 10% was not taken into account explicitly neither in the AG analysis nor in the admixture-corrected f4-symmetry tests also reported in Shinde et al. (2019).
What do they mean by deeply diverging WSHG ancestry? Divergent from which group, Iran Neo or IP?
Target: IRN_Shahr_I_Sokhta_BA2
Distance: 4.7397% / 0.04739683
57.4 IRN_Ganj_Dareh_N
34.2 S_AASI_Sim_Av
8.4 RUS_Tyumen_HG
0.0 NW_AASI_Sim_modern_Av
0.0 TUR_Marmara_Barcin_N
Target: IRN_Shahr_I_Sokhta_BA1
Distance: 3.5260% / 0.03526037
78.0 IRN_Ganj_Dareh_N
8.6 TUR_Marmara_Barcin_N
7.2 RUS_Tyumen_HG
3.6 NW_AASI_Sim_modern_Av
2.6 S_AASI_Sim_Av
^Sure you get 7-8% Tyumen ancestry when you use Ganj Dareh as the sole proxy. But adding in Hotu changes things:
Target: IRN_Shahr_I_Sokhta_BA2
Distance: 4.5444% / 0.04544424
34.2 S_AASI_Sim_Av
33.6 IRN_HotuIIIb_Meso
28.4 IRN_Ganj_Dareh_N
3.8 RUS_Tyumen_HG
0.0 NW_AASI_Sim_modern_Av
0.0 TUR_Marmara_Barcin_N
Target: IRN_Shahr_I_Sokhta_BA1
Distance: 3.3862% / 0.03386208
52.6 IRN_Ganj_Dareh_N
28.8 IRN_HotuIIIb_Meso
7.2 TUR_Marmara_Barcin_N
6.8 NW_AASI_Sim_modern_Av
2.8 RUS_Tyumen_HG
1.8 S_AASI_Sim_Av
^Now Tyumen goes down to 2-4%. Hotu is probably not ideal either, a hypothetical far eastern Iranian Hunter Gatherer would be better, which might reduce Tyumen to 0%. But even Hotu is enough to take the Tyumen ancestry down to less than half of the GD-only model value. I think the emphasis on 10% Tyumen is caused by using an ANE-poor Iranian source for the Indus Periphery samples.
Notice how Ganj Dareh does not vanish when Hotu as-a-proxy is introduced? This would imply multiple Iran-like contributions to Indus Periphery, going against both the stuff Shinde presented in 2019 (which precludes a later migration) and the standard “GD-only + Tyumen” models (which preclude older migrations). The model is probably more accurate this way.
One important thing to note is that the dental traits in Neolithic Mehrgarh are more like Mesolithic Uttar Pradesh and Chalcholithic Maharashtra than like Chalcholithic Mehrgarh. This would support a later migration between the neolithic and chalcolithic period. On the other hand, an early neolithic migration from zagros looks unlikely from the dental point of view.