Gene-level associations in suicide attempter families show overrepresentation of synaptic genes and genes differentially expressed in brain development.

Suicidal behavior (SB) has a complex etiology involving different polygenic and environmental components. Here we used an excess of significant markers (ESM) test to study gene-level associations in previous genome-wide association studies (GWAS) SNP data from a family-based sample, having medically severe suicide attempt (SA) as main outcome in the offspring. In SA without major psychiatric disorders (N = 498), a screening of 5,316 genes across the genome suggested association 17 genes (at fdr < 0.05). Genes RETREG1 (a.k.a. FAM134B), GSN, GNAS, and CACNA1D were particularly robust to different methodological variations. Comparison with the more widely used Multi-marker Analysis of GenoMic Annotation (MAGMA) methods, mainly supported RETREG1, GSN, RNASEH2B, UBE2H, and CACNA1D by using the "mean" model, and ranked 13 of the same genes as ESM among its top-17. Complementing the ESM screen by using MAGMA to analyze 17,899 genes, we observed excess of genes with p < .05 by using the "top" model, and the "mean" model suggested additional genes with genome-wide fdr < 0.25. Overrepresentation analysis of 10 selected gene sets using all genes with p < .05, showed significant results for synaptic genes, genes differentially expressed in brain development and for ~12% of the SA polygenic association genes identified previously in this sample. Exploratory analysis linked some of the ESM top-17 genes to psychotropic drugs and we examined the allelic heterogeneity in the previous SA candidate GRIN2B. This study complemented previous GWAS on SB outcomes, implicating both previous candidate (e.g., GRIN2B and GNAS) and novel genes in SA outcomes, as well as synaptic functions and brain development.