We then compared the results of GoShifter to those of the more commonly used matching-based enrichment tests2,5,7–10,12,13,19,29,31–38 (Table S1). These methods typically match SNPs on GEN, MAF, and TSS proximity.2,7,12,13,19,32,35 We observed that when we matched SNPs on these parameters, all simulated SNP sets, including 100% of those with functional variants derived from intergenic regions, obtained p < 0.05 for DHS enrichment (Figure 2A; Table S3). This suggests that selecting variants on GEN, MAF, and TSS proximity might be insufficient to control false-positive rates in the assessment of DHSs.