While associations to individual SNPs or genes may differ between specific items and their summed score, it is possible that similar gene-sets and pathways are implicated. To test this, we used the results of the gene-based analyses as input for MAGMA’s gene-set analysis23. For each of the 13 phenotypes we tested 7,244 gene-sets, comprising curated and gene ontology (GO) gene-sets derived from MsigDB24, and 53 tissue expression profiles25. The sum-score was significantly (P < 0.05/7,297 = 6.85 × 10−6) associated to 9 GO gene-sets, whereas 8 GO gene sets were associated to one or more individual items, but not to the sum-score (Supplementary Data 47; Fig. 2). Noteworthy, 4 out of 9 GO gene-sets that were found to be associated to the sum-score were also associated to individual items, indicating that the sum-score and the individual items at least in part implicate the same biological processes.
