A multi-omics study unravels the mechanism of water pollutants in gastric cancer: integrating network toxicology, machine learning, and tumor microenvironment remodeling.

Abstract

Water pollutants represent a growing environmental concern, yet their specific mechanisms in gastric cancer (GC) remain poorly understood. This study comprehensively investigates the multi-target mechanisms through which water pollutants promote gastric carcinogenesis using an integrated computational and bioinformatic approach. We screened 69 U.S. EPA-listed water contaminants for carcinogenicity using ADMETlab 3.0, ProTox-3, and IARC classifications, identifying seven high-risk pollutants. Their potential targets were predicted using five databases, and GC-related genes were identified from the GSE54129 dataset. Shared targets underwent functional enrichment, PPI network construction, and three machine learning algorithms to identify key targets. Diagnostic and prognostic analyses, immune infiltration, and single-cell sequencing explored tumor microenvironment remodeling, while molecular docking validated pollutant-target interactions. Results identified EGFR, MMP9, and CXCR4 as high-priority candidate key targets with significant diagnostic and prognostic value. These targets were implicated in cancer-related pathways and associated with immune cell infiltration. Molecular docking confirmed strong binding affinities between key pollutants and these targets. Our integrated analysis suggests that exposure to certain water pollutants may potentially contribute to gastric carcinogenesis through predicted interactions with EGFR, MMP9, and CXCR4, disrupting cancer-related signaling and remodeling the tumor microenvironment. These findings offer a computational framework for generating hypotheses regarding environmental risk assessment and may inform future investigations into therapeutic targets.