Gut microbiota has been implicated in various diseases, exerting systemic effects on the host by activating or inhibiting gene expression. However, the precise role of gut microbiota and its downstream targets in hepatocellular carcinoma (HCC) progression remains poorly understood. To address this gap, we developed a computational method, termed the microbial-associated signature (MAS) score, to characterize gut microbiota-induced transcriptional alterations across 33 cancer types and 29 normal tissues. MAS scores exhibited cancer type-specific associations with genomic alterations and immune microenvironment features, and potential clinical biomarker for immunotherapy responsiveness and recurrence in multiple malignancies. Single-cell MAS quantification revealed elevated microbiota-driven gene activation in tumor tissues relative to normal counterparts. Spatial transcriptomics further revealed enhanced microbe-mediated activation of key gene expression in hepatocellular carcinoma metastases. Using machine learning algorithms, we identified seven microbiota-regulated prognostic genes, with RHEB emerging as the top unfavorable predictor and subsequently validated experimentally. Our study provides a comprehensive framework for quantifying microbiota-driven transcriptional activity and its clinical implications in cancer.Copyright © 2025. Published by Elsevier B.V.