Table 5 . Three-population test in HGDP.

Source1	Source2	Target	f3	Z-score	αL	αU	ZSan	ZHan	αL > αR
Japanese	Italian	Uygur	−0.0259	−74.79	0.484	0.573	−46.08	−42.31
Japanese	Italian	Hazara	−0.0230	−74.05	0.46	0.615	−45.19	−42.22
Yoruba	Sardinian	Mozabite	−0.0211	−56.95	0.288	0.304	−40.65	−31.16
Mozabite	Surui	Maya	−0.0149	−19.67	0.165	0.408	−11.51	−9.40
Yoruba	San	Bantu-SA	−0.0107	−31.39	0.677	0.839	−24.67	−16.70
Yoruba	Sardinian	Palestinian	−0.0107	−36.70	0.07	0.157	−25.64	−18.35
Yoruba	Sardinian	Bedouin	−0.0104	−33.73	0.07	0.185	−23.37	−14.24
Druze	Yi	Burusho	−0.0090	−27.62	0.558	0.731	−15.94	−13.59
Sardinian	Karitiana	Russian	−0.0086	−20.68	0.694	0.923	−10.07	−10.98
Druze	Karitiana	Pathan	−0.0084	−22.25	0.547	0.922	−10.68	−9.37
Han	Orcadian	Tu	−0.0076	−20.64	0.875	0.926	−12.38	−8.98
Mbuti	Orcadian	Makrani	−0.0076	−19.56	0.038	0.151	−11.87	−6.61
Han	Orcadian	Mongola	−0.0075	−19.21	0.879	0.916	−12.63	−8.16
Han	French	Xibo	−0.0069	−16.92	0.888	0.922	−9.52	−8.19
Druze	Dai	Sindhi	−0.0067	−21.99	0.467	0.877	−12.25	−8.40
Sardinian	Karitiana	French	−0.0060	−18.36	0.816	0.964	−9.55	−9.33
Dai	Italian	Cambodian	−0.0060	−13.16	0.846	0.928	−6.78	−6.43
Sardinian	Karitiana	Adygei	−0.0057	−13.03	0.635	0.956	−5.60	−5.59
Biaka	Sardinian	Bantu-Kenya	−0.0054	−13.42	0.405	0.834	−9.65	−7.15
Sardinian	Karitiana	Tuscan	−0.0052	−11.26	0.803	0.962	−5.12	−4.76
Sardinian	Pima	Italian	−0.0045	−12.48	0.84	0.97	−7.48	−5.66
Druze	Karitiana	Balochi	−0.0044	−11.58	0.483	0.96	−6.96	−6.30
Daur	Dai	Han	−0.0026	−13.20	0.664	0.26	−7.89	−6.31	*
Han	Orcadian	Han-NChina	−0.0025	−7.09	0.958	0.97	−4.16	−2.74
Han	Yakut	Daur	−0.0025	−9.05	0.6	0.588	−6.91	−5.78	*
Druze	Karitiana	Brahui	−0.0025	−6.43	0.47	0.964	−2.23	−2.41
Hezhen	Dai	Tujia	−0.0021	−6.97	0.452	0.39	−4.36	−3.94	*
Sardinian	Karitiana	Orcadian	−0.0019	−4.31	0.803	0.952	−2.18	−3.24
She	Yakut	Oroqen	−0.0017	−5.13	0.422	0.296	−4.99	−2.44	*

This table only lists the most significantly negative f₃-statistics observed in HGDP samples. For each target population, we loop over all possible pairs of source populations, and report the pair that produces the most negative f₃-statistic. Here we only print results for target populations for which the most negative f₃-statistic is significant after correcting for multiple hypothesis testing; that is, the Z-score is more than 4 standard errors below zero. For the line with Bantu-SA as target, we used HGDP Han as an outgroup. In four cases indicated by an asterisk in the last column, the lower bound on the admixture proportion α_L is greater than the upper bound α_R, suggesting that our proposed three-population phylogeny is not feasible. We suspect that here the admixing (source) populations are themselves admixed. The 2 Z-score columns are with San and Han het ascertainment respectively.