Shared genetic architecture between schizophrenia and suicide attempt in European populations: fine-mapping and functional annotation
Y. DengIntroduction
Individuals with schizophrenia (SCZ) have an increased risk of suicide attempt (SA) and early mortality. Although previous studies have reported a genetic correlation between the two traits, the underlying mechanisms remain unclear. This study aims to investigate the shared genetic basis of SCZ and SA, and to identify potential causal variants and biological pathways.
Objectives
This study will (1) quantify the genetic architecture shared between SCZ and SA; (2) test whether the genetic overlap is better explained by a causal pathway or by pleiotropic effects; (3) identify genomic loci and causal variants that contribute to both traits; (4) characterize the biological pathways and brain regions enriched among the shared loci.
Methods
European-ancestry GWAS summary statistics from UKB, PGC and FinnGen (93,257 cases and 1,908,506 controls) were meta-analysed for SCZ and SA, respectively. SNP heritability and cross-trait genetic correlation were estimated via LD score regression. Mendelian randomisation and CAUSE distinguished causal from pleiotropic effects. Cross-trait meta-analysis using ASSET identified loci with shared association signals, followed by fine-mapping with SuSiE and regional colocalisation (coloc). Functionally relevant genes were prioritised using FUMA and MAGMA.
Results
The liability scale
h
2
was 0.118 for SCZ and 0.046 for SA. The genetic correlation between SCZ and SA was significant (r
g
= 0.42, SE = 0.03, p = 5.67e-54). MR analyses supported a positive causal effect of SCZ on SA (Table 1). However, CAUSE indicated that the causal model did not fit better than the sharing model significantly (p = .83). Fine-mapping prioritized 937 and 122 high-confidence credible sets (PIP > 0.5) across 192 and 28 risk loci for SCZ and SA, respectively. Among these, 20 loci contained credible sets in both traits, of which 6 showed strong evidence for shared causal variants (PP.H₄ > 0.9). These loci were mapped to genes including DRD2, ELAVL4, TSNARE1, which were also significant in gene-based test (Fig. 1). Tissue expression analysis showed no significant enrichment across GTEx tissues, although top signals were observed in cerebellum and cortical regions (Fig. 2). Gene-set analysis suggested significant enrichment of pathways related to synaptic structure and cellular stress regulation.
MR estimates (SCZ → SA) Table 1. Long description.