Cross-tissue transcriptome-wide association study prioritizes druggable targets for hepatocellular carcinoma

BackgroundHepatocellular carcinoma (HCC) is known as the sixth most common cancer and the third leading cause of death related to cancer. While genome-wide association study (GWAS) has uncovered risk loci, mapping variants to causal genes and tissues is challenging. The aim of this study is to translate HCC GWAS signals from variants to genes and tissues to identify genetically regulated causal candidates and potential therapeutic targets. MethodsWe collected and integrated 12 benchmark GWAS summary statistics from different research including various areas and causes of HCC. We applied S-PrediXcan using GTEx 49 tissues to infer genetically regulated expression. Multiple testing was controlled using group-wise Benjamini-Hochberg FDR within tissues (primary, q[≤]0.05). For visualization, suggestive signals with p[≤]0.01 were highlighted; heatmaps display FDR-thresholder results, whereas other figures show the full distribution. We then performed cross-tissue prioritization, gene-trait network construction, gene set enrichment analysis (GSEA), and pathway enrichment. ResultsWe identified 422 significant gene-tissue associations (BH-FDR within tissue, q[≤] 0.05) regarding HCC, consolidating to 34 unique genes. Top tissues included breast mammary, left ventricle, and esophagus muscularis. Excluding established HCC loci (e.g., PNPLA3, TM6SF2, TERT, HSD17B13), we prioritized 32 druggable "novel" candidates, with top-ranked examples including DHCR24 and HLA-DPB1. ConclusionsCross-tissue TWAS integrating GWAS and GTEx models identifies genetically regulated genes associated with HCC, refines tissue specificity, and prioritizes plausible drug targets. This gene-centric framework complements variant-level GWAS and supports mechanism-guided prevention and therapy development.