Table 4.

Lead gene-based population-specific results

Phenotype	Model	Gene	B38 positions	N	Variants	Singletons	P
Per90	HA	LINC01277	6:142,985,371–143,010,415	2803	2	0	5.02 × 10−8
		OR5K2	3:98,497,633–98,498,634	2803	1	0	2.74 × 10−7
	AA females	S100A16*	1:153,607,528–153,616,353	1009	1	1	2.07 × 10−7
		CSMD2-AS1	1:33,867,977–33,885,456	1009	1	1	2.07 × 10−7
	EA females	MRPS33	7:141,006,422–141,014,911	1702	9	8	1.22 × 10−9
		LINC01811	3:34,170,921–34,558,474	1702	6	5	9.71 × 10−8
		NELFCD*	20:58,980,722–58,995,761	1702	12	10	3.32 × 10−7
		SLC22A8*	11:62,988,399–63,015,986	1702	3	3	3.58 × 10−7
	HA females	AL132709.1	14:101,077,452–101,077,578	1660	2	0	1.41 × 10−7
		EPHX4	1:92,029,443–92,063,474	1660	12	10	3.48 × 10−7
	HA males	C16orf90	16:3,493,483–3,496,479	1143	6	3	1.36 × 10−8
		TVP23B	17:18,781,270–18,806,714	1143	4	4	2.53 × 10−7
		IPCEF1	6:154,154,536–154,356,890	1143	10	8	4.07 × 10−7

Lead MMSKAT gene-based population-specific associations within one order of magnitude of significance (p < 4.51 × 10⁻⁸) are shown. The Variants column indicates the number of filtered polymorphic variants with minor allele frequency < 5% available for testing, a portion of which were singletons. *Druggable gene [59, 60]. Full results for genes with p < 0.01, including descriptions, are provided in Additional file 1: Table S5. A list of individual variants comprising each gene is provided in Additional file 1: Table S6