Integrating bulk and single cell sequencing data to identify prognostic biomarkers and drug candidates in HBV associated hepatocellular carcinoma.

Hepatitis B virus (HBV) infection is a major driver of hepatocellular carcinoma (HCC), yet the mechanisms by which HBV triggers HCC and how it interacts with the immune system remain largely undefined. In this study, 53 immune-related key genes involved in HBV-associated HCC progression were identified. By analyzing the mean C-index of 101 machine learning models, the optimal model-combining stepwise Cox regression (forward) with RSF-was developed to characterize the immune risk index. Patients in the high-risk group exhibited worse survival outcomes and increased infiltration of immunosuppressive cells. Integrating PPI analysis with machine learning, SPP1, GHR, and ESR1 emerged as promising druggable targets, with SPP1 notably overexpressed in tumors and linked to adverse outcomes. ScRNA-seq analysis revealed SPP1 was predominantly expressed in angio-TAMs, which may impair anti-tumor immunity by limiting T and NK cell infiltration. It also involved in tumor progression via angiogenesis and EMT pathways. Drug prediction and molecular docking identified small molecules such as myricetin and mefloquine that can target the aforementioned key immune genes, thereby modulating the immune landscape of HBV-HCC. Repurposing these established drugs represents a novel therapeutic avenue, offering both efficacy and expedited clinical translation for HBV-HCC.