Table 2.

Three novel independent SNPs significantly associated with BCC risk in both GWAS studies and the meta-analysis.

SNP	Region	DNA repair gene	Major/minor	MAF	Meta-analysis		23andMe		Harvard		Phet	I2
					OR	P	OR	P	OR	P
rs2805831	9q22.33	XPA	G/A	0.17	0.93	1.35×10−6	0.93	3.00×10−5	0.92	1.22×10−2	0.75	0.0%
rs659857	11q13.1	MUS81	C/T	0.33	1.06	3.09×10−6	1.06	2.35×10−5	1.05	5.00×10−2	0.83	0.0%
rs57343616	12q13.3	NABP2	C/T	0.07	1.11	6.47×10−6	1.11	1.36×10−4	1.13	1.81×10−2	0.72	0.0%

SNPs located in novel BCC susceptibility loci that met significance level after multiple testing corrections (P < 1.69 × 10⁻⁵) in the overall meta-analysis and also exhibited significant associations (P < 0.05) in both GWAS datasets with the same effect direction are listed. We used minor alleles as effect alleles in the analysis. Additionally, we report genetic locus, related DNA repair genes, major allele, minor allele, minor allele frequency (MAF) as calculated from overall meta-analysis, and odds ratio (OR) with P value for each GWAS, calculated with respect to the minor allele. In 23andMe GWAS, we analyzed 12,945 self-reported BCC cases and 274,252 controls of European ancestry in the US. In Harvard GWAS, we analyzed 4,242 self-reported BCC cases and 12,802 controls of European ancestry in the US. We then combined the data from these 2 GWAS studies (which resulted in 17,187 cases and 287,054 controls) and performed fixed-effect meta-analysis. Statistics for effect heterogeneity were shown by P_het and I².