Table 1.

Replication analysis of novel candidate GV susceptibility loci

		GWA Study				Replication 1		Replication 2		Meta-analysis Replication 1 + Replication 2		Meta-analysis GWA + Replication 1 + Replication 2
SNP	Risk allele	AFRA, cases	AFRA, controls	PLINK P	OR	P	OR	P	OR	P	OR	P	OR
Replicated
rs17008723	G	0.214	0.171	3.65 × 10-6	1.32	1.47 × 10-3	1.35	0.017	1.26	1.36 × 10-3	1.36	1.04 × 10-8	1.33
rs6902119	C	0.495	0.446	5.72 × 10-5	1.21	2.92 × 10-3	1.28	0.058	1.09	3.79 × 10-3	1.25	3.94 × 10-7	1.23
Uncertain
rs2603127	A	0.247	0.213	2.69 × 10-7	1.34	0.404	1.09	0.078	1.00	0.085	1.19	5.36 × 10-8	1.30

SNPs rs17008723, rs2603127, and rs6902119 were located at nt 71,573,135 in the FOXP1 region of 3p13; nt 167,505,791 in the CCR6 region of 6q27; and nt 108,243,551 in the MYH15 region of 3q13.13, respectively. SNP nucleotide positions are from GRCh37 and genes in close proximity to the designated SNP are denoted. AF_RA, allele frequency of the risk allele. EIGENSTRAT GWA P-values for SNPs rs17008723, rs6902119, and rs2603127 were 6.25 × 10^-6, 7.23 × 10^-5, and 4.50 × 10^-7, respectively. PLINK and EIGENSTRAT GWA test statistics were calculated and adjusted for the genomic inflation factor 1.048 as described in the Supplementary Methods. The Bonferroni adjusted significance threshold for the combined replication stage 1 + 2 meta-analysis was P < 1.00 × 10^-2, and the significance threshold for the overall combined GWA + replication stage 1 + 2 meta-analysis was P < 5 × 10^-8. FOXP1 SNP rs17008713 from the GWA study could not be genotyped in the replication study for technical reasons; therefore, genotypes for nearby SNP rs17008723 (r² = 0.99) were imputed, and rs17008723 was genotyped in the replication study. Data for SNPs rs2192346 and rs7868451, for which association was excluded, are shown in Supplementary Table 3.