Integrative machine learning and network toxicology framework for assessing environmental pollutant TCDD-induced osteoarthritis risk: a computational cheminformatics approach.

Objective: Osteoarthritis (OA) is a prevalent degenerative joint disease influenced by both genetic susceptibility and environmental factors. Increasing evidence suggests that exposure to persistent environmental pollutants, such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), may contribute to OA progression; however, the molecular mechanisms underlying these effects remain unclear. Therefore, this study aimed to elucidate the potential molecular interactions and mechanisms by which TCDD may influence OA development using an integrated systems-level computational strategy. Methods: OA-related genes were identified by integrating multiple transcriptomic datasets from the Gene Expression Omnibus (GEO) database through differential expression analysis and weighted gene co-expression network analysis (WGCNA). Putative TCDD targets were collected from public chemical-protein interaction databases, and overlapping targets were identified to construct a disease-toxicant interaction network. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed to explore the biological functions and pathways involved. Key OA-related proteins associated with TCDD exposure were further prioritized using an integrative machine learning framework. Immune cell infiltration analysis was conducted to investigate associations between hub proteins and the OA immune microenvironment. Finally, molecular docking and molecular dynamics simulations were employed to evaluate the stability and binding behavior of TCDD with the identified key proteins. Results: A total of 471 OA-related genes were identified, among which 43 overlapped with predicted TCDD targets. Functional enrichment analysis indicated that these targets were primarily involved in inflammatory signaling, arachidonic acid metabolism, calcium signaling, and G protein-coupled receptor-related pathways. Machine learning analysis identified four hub proteins-PTGS1, CBR1, HTR2B, and PTGS2-with robust diagnostic relevance across independent cohorts. Immune infiltration analysis revealed that these hub proteins were significantly associated with macrophage polarization, mast cell activation, and T-cell dysregulation in OA. Molecular docking and molecular dynamics simulations demonstrated stable binding conformations between TCDD and all four hub proteins, with trajectory analyses confirming persistent ligand binding and structural stability throughout the simulations. Conclusion: PTGS1, CBR1, HTR2B, and PTGS2 were identified as key hub proteins potentially mediating TCDD-associated osteoarthritis. Our integrated computational framework highlights specific inflammatory, metabolic, and neurotransmitter pathways linking environmental pollutant exposure to OA pathogenesis, providing actionable targets for future experimental validation.