Table 2.

Comparison of Signals of Population Differentiation

SNP ID	Region	Gene	Jin et al.9data			Bhatia et al.23Data
			WC FST	Hudson FST	Model-Based p Value	Model-Based p Value
rs1541044	chr1: 100,125,058–100,183,875	–	0.0562	0.0439a	4.7 × 10−5	0.04
rs4460629	chr1: 153,401,959–153,464,086	–	0.0692	0.0650	6.8 × 10−7	2.1 × 10−4
rs12094201	chr1: 236,509,336	–	0.0561	0.0489	1.7 × 10−5	0.86
rs7642575	chr3: 31,400,165	–	0.0453	0.0393a	1.1 × 10−4	0.41
rs652888	chr6: 26,554,684–33,961,049	HLA	0.0711	0.0627	1.1 × 10−6	1.8 × 10−11
rs9478984	chr6: 151,555,551–151,569,258	–	0.0545	0.0596	2.1 × 10−6	0.02
rs10499542	chr7: 22,235,870	–	0.0461	0.0453	3.6 × 10−5	0.35
rs304735	chr7: 79,768,487–80,482,597	CD36	0.0946	0.0690	3.0 × 10−7	3.7 × 10−13
rs2920283	chr8: 143,754,039–143,758,933	PSCA	0.0468	0.0532	7.6 × 10−6	6.4 × 10−7
rs1498487	chr11: 5,034,229–5,421,456	HBB	0.0617	0.0464	2.4 × 10−5	1.7 × 10−7
rs4883422	chr12: 7,189,594	–	0.0472	0.0461	3.0 × 10−5	1.3 × 10−3
rs6491096	chr13: 25,488,362	–	0.0472	0.0373a	1.5 × 10−4	0.4
rs1075875	chr16: 47,595,721	–	0.0766	0.0608	1.3 × 10−6	NAb
rs6015945	chr20: 59,319,574	–	0.0627	0.0550	4.3 × 10−6	0.5

We recreated Table 2 from Jin et al.⁹ by analyzing the same data with the Hudson instead of the WC estimator. We also estimated the p value at each SNP by using the reported F_ST = 0.0007 of Jin et al.⁹ and a model-based approach.²⁴ Finally, we report the model-based p value of the most significant SNP in the region from the parallel study by Bhatia et al.²³ We note that results reported in that paper were more significant than those reported here because Bhatia et al. analyzed additional populations. All positions are from UCSC Genome Browser build hg18.

^aThese loci fell below the threshold for the 99.99^th percentile (0.0452) when the Hudson estimator was used.

^bThis locus was not available (NA) because it lacked data in the Bhatia et al.²³ study.