Table 3.

Lookup of previous GWAS findings.

SNP	chr	position	gene	EA	OA	EAF	HIGH			LOW/MID
							beta	se	P	beta	se	P
rs11928865	3	7155702	GRM7	A	T	0.260	−0.0058	0.0170	0.7313	−0.0256	0.0160	0.1100
rs779706	3	7524042	GRM7	C	G	0.325	−0.0050	0.0180	0.7806	−0.0043	0.0171	0.7998
rs779701	3	7518772	GRM7	a	g	0.688	0.0038	0.0180	0.8322	0.0044	0.0169	0.7956
rs457717	5	75920972	IQGAP2	a	g	0.359	0.0079	0.0156	0.6151	0.0193	0.0148	0.1927
rs681524	11	116748314	SIK3	t	c	0.930	−0.0238	0.0307	0.4376	−0.0122	0.0289	0.6716
rs4932196	15	89253268	ISG20	t	c	0.803	0.0682	0.0192	0.0004	0.0841	0.0181	3.44E-06
rs5756795	22	38122122	TRIOBP	t	c	0.556	−0.0228	0.0148	0.1235	−0.0662	0.0141	2.61E-06
rs2877561	3	121712051	ILDR1	a	c	0.288	0.0986	0.0162	1.26E-09	0.0676	0.0155	1.29E-05
rs9493627	6	133789728	EYA4	a	g	0.338	0.0173	0.0155	0.2650	0.0402	0.0146	0.0060

All significant SNPs from previous GWA studies on ARHI were included. SNPs in ILDR1, ISG20 and TRIOBP showed significant replication in our discovery cohort (P < 0.0056, Bonferroni corrected, in bold). EYA4 was just above significance level. For GRM7, the phenotype consisted of Z-scores of frequencies 2, 4 and 8 kHz, normalized to means derived from the ISO standard²⁷. Phenotyping for IQGAP2 and SIK3 was done through principal components analysis of audiometry^24,26, while EYA4, ILDR1, ISG20 and TRIOBP were found by using ICD-9 diagnoses of hearing loss⁶⁵.