Transcriptomic Identification of Diagnostic Biomarkers for Alcohol-Associated Liver Cirrhosis: Integration of Population-Level Epidemiology with Multi-Cohort Transcriptomic Analysis

Alcohol-associated liver cirrhosis (ALC) lacks aetiology-specific molecular diagnostic biomarkers. This study aims to quantify the association between alcohol and cirrhosis risk, and to identify transcriptomic diagnostic biomarkers and candidate therapeutics. Methods: Survey-weighted logistic regression was applied to 17,007 adults from NHANES (2017–2023) to quantify alcohol-cirrhosis associations. ALC transcriptomic data from four GEO datasets were analysed using weighted gene co-expression network analysis (WGCNA) and three parallel machine learning algorithms (LASSO, Random Forest, SVM-RFE). External validation was performed in an independent cohort of 93 samples. Candidate therapeutics were identified via drug signature database querying and validated by molecular docking. Heavy drinking conferred a 5.14-fold increased cirrhosis risk (95% CI: 2.60–10.20, p < 0.001). Transcriptomic analysis revealed global downregulation of long non-coding RNAs (with 91.7% of dysregulated lncRNAs being suppressed). A five-gene diagnostic signature (IL1B, CCL3, LUM, SPP1, ITGA6), specifically developed to distinguish ALC from histologically normal liver tissue, achieved an area under the receiver operating characteristic curve (AUC) of 0.824 in an external validation cohort. Immune infiltration analysis uncovered global contraction of macrophage-associated transcriptomic signatures across M0, M1, and M2 subtypes, inversely correlated with fibrotic hub gene upregulation. Fluvastatin and honokiol were identified as candidate therapeutic agents, with strong binding affinities to IL1B and CCL3, respectively. This study confirms a dose-dependent alcohol-cirrhosis association and establishes a five-gene diagnostic signature (distinguishing ALC from normal liver tissue) alongside candidate therapeutics, warranting prospective clinical validation. The identified tissue-derived signature and therapeutic candidates provide a foundation for future ALC-specific diagnostic and therapeutic strategies; their translation into a non-invasive (e.g., blood-based) assay will require dedicated validation in circulating samples.