It is well documented that a SNP overlapping with a functional region is more likely to be a functional SNP [41]. To identify the causal SNPs in each identified gene, we used the annotation data from RegulomeDB [42], and performed a fine-mapping analysis called Probabilistic Identification of Causal SNPs (PICS) [43] to estimate the possibility of each SNP in the locus to be a causal SNP. By using information derived from the permutations, PICS score was calculated as the posterior probability of each SNP being the causal variant, given the observed pattern of association at the locus [43].
