Genomic structural equation modeling uncovers shared genetic architecture and comorbidity mechanisms of lung function decline.
TL;DR
The mvLung GWAS characterise shared genetic signals across the included traits and nominate candidates for functional follow-up and drug-target analyses showed pathway-level concordance with previously studied cardiopulmonary drug classes.
Abstract
Chronic respiratory diseases cause substantial global morbidity and mortality, yet the shared genetic architecture of pulmonary-function traits, exposure-related phenotypes, and cardiometabolic traits remains incompletely characterised. We applied genomic structural equation modelling to GWAS summary statistics for four pulmonary-function traits, lifetime smoking index, ambient PM2.5 exposure, and arterial oxygen tension to model their shared genetic covariance (mvLung). The mvLung GWAS identified 2,156 loci, including 376 not significant in the individual input GWASs. EFEMP1 was prioritised as a candidate gene, and 41 respiratory and cardiometabolic traits showed significant genetic correlations. Colocalisation, gene-prioritisation, and enrichment analyses identified candidate shared regions and highlighted extracellular-matrix, cell-adhesion, epithelial-barrier, and signalling-related annotations. Drug-target analyses showed pathway-level concordance with previously studied cardiopulmonary drug classes. These findings characterise shared genetic signals across the included traits and nominate candidates for functional follow-up. Because mvLung incorporates exposure-related and oxygenation phenotypes, the results do not represent direct genetic effects on pulmonary function alone, causal mechanisms, or therapeutic efficacy.