Unraveling the Carcinogenic Mechanisms of Food Contaminants: An Integrated in Silico Framework Combining Network Toxicology, Machine Learning, and Molecular Docking

Food contamination poses a significant global health threat with carcinogenic potential, though the molecular pathways connecting contaminants to cancer remain poorly understood. This study sought to identify key molecular targets mediating the carcinogenic effects of nine prevalent dietary contaminants: glyphosate, perfluorooctane sulfonate, nitrosamines, pentabromodiphenyl ethers, methylmercury, dioxins, acrylamide, pyrrolizidine alkaloids, and aflatoxin. Using multiple online databases, we identified target genes associated with these contaminants and pan-cancer, then conducted protein-protein interaction (PPI) analysis and visualization on intersecting genes. Subsequent gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) functional enrichment analyses were performed to uncover potential mechanisms, with a focus on breast (BRCA), prostate (PRAD), and colon (COAD) carcinomas due to their significant pathway associations. Hub genes were prioritized through an integrative strategy combining topological algorithms in cytoscape (Centiscape, MCODE, and cytohubba's MCC), machine learning validation, and weighted gene co-expression network analysis (WGCNA). Molecular docking simulations were conducted to examine interactions between contaminants and hub genes. The study identified 69 pan-cancer-intersected targets, with enrichment analyses revealing significant cancer-associated pathways. Hub gene prioritization pinpointed JUN in BRCA, CDC42 in COAD, and MAPK14 in PRAD as critical regulatory targets. Validation using The Cancer Genome Atlas (TCGA) data confirmed statistically significant differential expression patterns (p < 0.05) for these targets across respective malignancies. Gene set enrichment analysis (GSEA) outlined pathway activation profiles consistent with tumor progression mechanisms. Molecular docking simulations demonstrated strong binding affinities (binding energy ≤ -5.0 kcal/mol) between contaminants and structural domains of the identified hub targets, suggesting potential mechanistic links between these food contaminants and cancer development.