Decoding the shared genetic architecture of atherosclerotic cardiovascular disease: a genomic structural equation modeling analysis
摘要
Atherosclerotic cardiovascular disease (ASCVD) represents a broad spectrum of phenotypes with shared pathology. However, the joint genetic architecture across different vascular beds remains incompletely characterized.
MethodsWe applied genomic structural equation modeling (Genomic SEM) to integrate GWAS summary statistics from six ASCVD-related traits: coronary artery disease, myocardial infarction, ischemic stroke, peripheral artery disease, carotid intima-media thickness, and intracranial aneurysm. Downstream analyses included multivariate GWAS, fine-mapping, TWAS with FOCUS, MAGMA, pathway enrichment, cell-type and functional annotation, gsMAP and polygenic risk score.
ResultsA single latent common factor model demonstrated an excellent fit to the genetic covariance matrix (CFI = 0.9837, SRMR = 0.1448). The multivariate GWAS identified 839 genome-wide significant SNPs, delineating 85 independent lead loci, of which 67 were novel. Integration of TWAS and FOCUS prioritized 6 high-confidence putative effector genes, including CDKN2B, MIA3, and NBEAL1. Functional enrichment highlighted lipid remodeling and apolipoprotein binding as predominant pathways. Notably, heritability was significantly localized in endothelial cell lineages across multiple tissues. Spatial mapping further identified the lung and kidney as key developmental tissue contexts for ASCVD genetic risk.
ConclusionOur study provides a unified genetic framework for ASCVD, demonstrating that its shared liability is anchored in lipid homeostasis and systemic endothelial dysfunction. These findings offer a foundational atlas for early genomic-based risk stratification and therapeutic targeting.