Answers are grounded in this paper and cite their sources.
Hiatus hernia (HH), gastroesophageal reflux disease (GERD), Barrett esophagus (BE), and esophageal adenocarcinoma (EAC) are clinically related esophageal phenotypes, but their joint genetic architecture has not been characterized within a unified multivariate framework. We applied genomic structural equation modelling (Genomic SEM) to summary statistics from genome-wide association studies (GWAS) of HH, GERD, BE, and EAC. A single common factor was constructed based on the cross-trait genetic covariance matrix estimated by linkage disequilibrium score regression, and factor-level multivariate GWAS was performed to identify shared genetic variants. Independent significant SNPs and lead SNPs were defined using FUMA, followed by statistical fine-mapping through SuSiE and FINEMAP to prioritize candidate variants. Effector genes were prioritized by combining MAGMA gene-based analysis with sCCA-TWAS and FOCUS, and the biological context was characterized through pathway enrichment, cell-type enrichment, stratified heritability partitioning, and spatial tissue mapping. The single common factor model yielded a CFI of 0.947 and an SRMR of 0.119, and all four phenotypes loaded positively and significantly on the factor. Factor-level analysis identified 370 genome-wide significant variants, which FUMA classified into 277 lead SNPs under the prespecified LD criteria. Fine-mapping identified high-confidence candidate variants at loci including FOXP4 and CDKN2B-AS1. Gene prioritization established the known EAC susceptibility gene CRTC1 and the lipid raft scaffolding protein FLOT1 as candidate effector genes. Heritability was enriched in broadly defined H3K4me1-marked enhancer regions, whereas no tissue- or cell-type-specific annotation remained significant after multiple-testing correction. This study provides an integrative multivariate characterization of shared genetic liability across four clinically related esophageal phenotypes and identifies candidate regulatory signals for further investigation.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: