Deciphering the role of microplastics-related molecular signatures in cervical cancer progression: Transcriptomic analysis and diagnostic model development.

Environmental microplastic exposure has been increasingly implicated in cervical carcinogenesis, yet the underlying biological mechanisms of microplastic-associated disease progression remain largely unexplored. Therefore, this study combined transcriptomic analysis, chemical-gene interaction data, machine learning, and molecular docking to investigate the molecular impact of microplastics exposure associated with disease progression. Transcriptomic data from GEO database was retrieved and differential expression analysis was performed, yielding a consensus core of 330 common genes across disease conditions. Microplastic candidates including Polystyrene, Polyethylene, Polypropylene, Nylon 612, Polymethyl Methacrylate, and Polyvinyl Chloride were selected based on established in vitro and in vivo evidence of carcinogenic potential in cervical tissues. Cross-referencing with chemical-gene interaction data identified 15 candidate genes linked to microplastic exposure and cervical cancer. The combinatorial feature selection across all 15-gene combinations (n = 32,570) using Support Vector Machine and Multinomial Logistic Regression algorithms yielded an optimized 5-gene diagnostic signature (FN1, HTR4, KIF14, PTP4A3, SLC7A2) for multiclass classification of Normal, CIN, and Cervical Cancer samples. The optimized SVM model achieved multiclass AUC values of 0.937, 0.889, and 0.925 for the Normal, cervical intraepithelial neoplasia (CIN), and Cancer classes, respectively, with further validation in independent GSE27678 validation cohort, demonstrating its robustness and generalizability. Molecular docking using AutoDock Vina demonstrated that polystyrene exhibited the highest binding affinities across all five target proteins, with peak scores of -6.681 kcal/mol, with KIF14. Collectively, these findings highlight potential molecular interactions between microplastics, particularly polystyrene, and key genes involved in cervical cancer progression. Thus, the proposed five-gene provides a promising framework for multiclass diagnosis and future studies of microplastics-associated carcinogenesis.