Gene-based association tests identified seven genes as significant after Bonferroni correction for testing >16,000 genes: MAGI2 (p<1.0×10−6), NAV2 (p<1.0×10−6), CACNA1C (p=1.0×10−6), PCDH9 (p=1.0×10−6), MYO16 (p=1.0×10−6), IQCH (p=2.0×10−6), DLGAP1 (p<1.0×10−6). We examined overlap of these novel “candidate” genes derived from the CADD GWAS with results from MCTFR and SAGE as a single “pathway” (i.e., gene set) in INRICH (Lee 2012). This allowed us to estimate whether specific genes identified in the CADD results overlapped with the low p-value genomic regions (i.e., loci tagged at r2>0.5 by a SNP reaching GWAS p<5×10−3) in the MCTFR and SAGE results more than expected by chance. The CADD-identified gene set was not significant in analysis of either the MCTFR (0 regions overlapped genes identified in CADD, p=1.0) or SAGE samples (6 regions overlapped genes identified in CADD, p=0.14).
