Artificial intelligence and multi-omics revealed the authentic regulation of ADP-ribosylation in modulation Treg/Th17 ecosystem and multi-target therapeutic strategy enrichment for HBV+ Hepatocellular Carcinoma patients

Dysregulation of Protein post-translation(PTM) and Treg/Th17 balance contribute to the progression of Hepatocellular Carcinoma(HCC) patients. Our study aims to investigate the ADP-ribosylation mechanisms in regulation of Treg/Th17 ecosystem for HBV + Hepatocellular Carcinoma patients. We first acquired ADP-ribosylation and Treg/Th17 ecosystem(AT)-associated shared differentially expressed genes(sDEGs) from public bulk profiles of HBV + Hepatocellular Carcinoma(H-HCC) patients via integrative bioinformatic pipelines(Limma, ssGSEA and WGCNA). Next, Cox regression and integrated machine learning framework identified AT-associated risk groups and hub genes for H-HCC patients. In addition, characters of hub genes in H-HCC were deciphered at bulk and single-cell levels of H-HCC patients. Indeed, ridge regression and artificial intelligence(AI) pipeline(DrugRefLector) enabled the identification of therapeutic agents for AT-associated risk group and H-HCC patients. Finally, in vitro assays estimated the Treg-distributed hub gene pathogenic role in HCC. AT can conduct the risk stratification of H-HCC patients, and PARP1, KPNB1 and FARP2 can be considered as AT-associated hub modulator involved in H-HCC pathogenesis, which was mainly distributed in Treg. Besides, therapeutic agents enriched by machine learning and deep learning pipelines can potentially elaborate drug synergy effects for H-HCC patients. This study first elucidated the AT mechanism and corresponding predictive and therapeutic potentials for H-HCC patients.