To address these limitations, we propose a new SNP-based pathway analysis method, called SNP Set Enrichment Analysis (SSEA), for GWAS studies. SSEA consists of two main steps: selecting representative SNPs for each gene, and performing pathway enrichment analysis using all selected SNPs. In the first step, we exploit an adaptive rank truncated product method with permutations to choose the most significant subset of SNPs for each gene. The number of SNPs representing a gene is not predetermined, but data driven. Then for each pathway, we calculate the average number of representative SNPs for the genes within this pathway and re-select SNPs using this number. In the second step, we modify the existing GSEA algorithm [17] to conduct the pathway enrichment analysis using all selected SNPs. We rank all SNPs selected from the first step based on their strength of association with the trait, and then test whether the set of SNPs associated within a pathway is significantly enriched with high ranks using a weighted Kolmogorov-Smirnov test. Because this test is rank-based, SNPs with smaller P-values tend to contribute more in a pathway.
