The scoring metric used here is as follows: For each gene g, the minimum GWA p-value of all SNPs with index i that fall within the extended gene boundaries (see Step 1) is chosen, :(1)where I(g) is the set of indeces of SNPs whose chromosome positions fall between the extended gene boundaries. is the association p-value for SNP i calculated in a GWA study or meta-analysis (see GWA studies and meta-analyses section). A z-score is then computed based on for each gene g, using a mean of 0 and standard deviation of 1, assuming a normal distribution. should be most powerful for genes that contain one major target region or haplotype with potential causal mutations in or around their coding sequence. Other SNP to gene scoring metrics can be used here.
