AI-Driven Multiscale Study on the Mechanism of Polygonati Rhizoma in Regulating Immune Function in STAD.

Polygonati Rhizoma, a traditional Chinese medicine, has demonstrated immunomodulatory and anticancer properties, yet its precise mechanisms in stomach adenocarcinoma (STAD) remain underexplored. This study aims to uncover the multitarget mechanisms of Polygonati Rhizoma in regulating the tumor immune microenvironment in STAD using artificial intelligence (AI)-driven network pharmacology, bioinformatics, and single-cell RNA sequencing, offering new insights into its immunotherapeutic potential. This study harnessed the power of AI to unravel the molecular mechanisms underlying Polygonati Rhizoma's effects. AI-driven methodologies screened 38 putative constituents, retaining 8 based on ADME criteria. Machine Learning algorithms predicted potential targets, which were cross-referenced with 5,569 immune-related genes from GeneCards, revealing 52 immune-associated targets. Differential expression analysis of the STAD data set identified 18 overlapping DEGs with prognostic significance and immune cell infiltration correlations. Key targets (AKT1, TP53, PTGS2 and VEGFA) emerged as central nodes in the network, with AI-assisted molecular docking confirming strong binding affinities, particularly between diosgenin and these core proteins. Molecular dynamics simulations further validated these interactions. Single-cell RNA sequencing revealed distinct target-gene expression patterns across malignant, stromal, and immune cell subsets in digestive-system tumors. In vitro, Polygonati Rhizoma extract significantly inhibited HGC-27 cell viability and increased intracellular ROS levels. These findings underscore the critical role of AI in integrating multiscale analyses, unveiling a multitarget immunomodulatory and antitumor mechanism for Polygonati Rhizoma in STAD, and providing a foundation for future preclinical and clinical studies.