Among extant competitive enrichment methods, three analytical frameworks predominate. In the first of these, threshold-based approaches, hypergeometric, chi-square, or Fisher’s exact test statistics are used to identify pathways that are overrepresented among the “significant” markers under study. Notably, the threshold for “significance” is arbitrary and can affect results [48]; observed SNP-level thresholds have ranged from p < 0.05 [37] to p < 5 × 10−8 [34]. In contrast, rank-based approaches order all of the markers being studied by their significance and then test for pathways that have lower rankings than the overall distribution. While the rank-based tools GenGen [49] and GSEA-SNP [50] use a Kolmogorov-Smirnov-like running sum that gives greater weight to more significant markers, others rely on MAXMEAN-related statistics as potentially powerful and efficient alternatives [51–53]. Compared with threshold-based methods, rank-based approaches more naturally account for differences in significance among markers [24] but may also be heavily influenced by a few highly-significant markers [54]. Finally, z-score methods infer enrichment based on deviation from a normal distribution that accounts for the size of each pathway [52, 55]; while these methods
