Chunk #59 — Materials and Methods — Meta-Analysis Gene-set Enrichment of variaNT Associations (MAGENTA) — Step 4: Gene set enrichment analysis of genome-wide association data
retained. This was done to eliminate potential inflation of gene set enrichment significance due to two or more genes in a gene set that are physically proximal along the chromosome and hence may capture the same association signal (assuming one causal gene per associated locus). This yielded an effective number and set of genes that was used for the next steps of the GSEA test. (iii) For each gene set gs the fraction of genes with < was recorded (denoted here as the ‘leading edge fraction’), where is a predetermined gene p-value cutoff, defined as a given percentile of all gene p-values in the genome. is specific for a given GWA study or meta-analysis. In this study, we used = 95th percentile of for all genes g in the genome, as it gave the optimal power of five cutoffs tested (99th, 95th, 90th, 75th, and 50th percentile of all gene p-values) with power simulations (see Figure S5 and Simulations section below). (iv) Finally, a nominal GSEA p-value, was calculated for each gene set gs, defined as the fraction of randomly sampled gene sets of identical set size, whose leading edge fraction is equivalent to or larger than the observed leading
