Transcriptome and plasma proteome analyses identify susceptibility genes and proteins for alcohol-related...

Genome-wide association studies (GWAS) have identified loci associated with alcohol-related hepatocellular carcinoma (ALD-HCC); however, most risk variants lie in noncoding regions and the downstream regulatory architecture remains incompletely characterized. Integrating genetically predicted transcriptomic and circulating proteomic signals may facilitate target prioritization and biomarker discovery.

An integrative analytic framework combining transcriptome-wide association study (TWAS), cross-tissue sparse canonical correlation analysis with the aggregate Cauchy association test (sCCA+ACAT), and plasma proteome-wide association study (PWAS) was applied to ALD-HCC GWAS summary statistics from 2,107 unrelated European individuals with alcohol-related liver disease (775 cases and 1,332 controls; 7,962,325 variants). Conditional and bivariate analyses were conducted to assess independence and potential shared cis-genetic regulation between transcriptomic and proteomic signals at risk loci. Colocalization analysis was further conducted to assess whether the identified signal pairs shared a causal genetic variant.

Fifteen previously unreported susceptibility genes were prioritized (11 by TWAS and 4 by sCCA+ACAT), and PWAS identified NCAN as a significant plasma protein-associated signal at an ALD-HCC risk locus. Conditional and bivariate analyses supported associations with genetically predicted expression of nine genes and genetically predicted plasma NCAN levels, indicating convergent transcript-protein signals within established risk regions. Through colocalization analysis in GWAS loci, we additionally identified 5 putative susceptibility genes.

Integrative TWAS/sCCA+ACAT/PWAS analyses nominate candidate susceptibility genes and identify NCAN as a plausible circulating locus-level signal relevant to ALD-HCC risk, providing a genetic resource for further mechanistic investigation.