Unraveling the molecular mechanisms of paclitaxel in high-grade serous ovarian cancer through network pharmacology.

High-grade serous ovarian cancer (HGSOC) is the most common and aggressive subtype of epithelial ovarian cancer, often diagnosed at advanced stages with a poor prognosis. Paclitaxel (PTX), a standard chemotherapeutic agent for HGSOC, exerts cytotoxic effects on cancer cells and modulates the tumor microenvironment. This study aimed to elucidate the molecular mechanisms of PTX in HGSOC using bioinformatics, machine learning, network pharmacology, and molecular docking, to identify potential diagnostic biomarkers and therapeutic targets. We identified differentially expressed genes (DEGs) between HGSOC and normal ovarian tissues using the GSE54388 dataset from the Gene Expression Omnibus database. The intersection of these DEGs with PTX targets, identified from the Swiss Target Prediction database, yielded 15 overlapping genes. These genes were analyzed via protein-protein interaction (PPI) network analysis to identify significant interaction relationships. Kaplan-Meier survival analysis was then performed to assess the prognostic significance of these genes. Their protein expression patterns in HGSOC tissues were validated using the Human Protein Atlas (HPA) database. Functional enrichment analysis was conducted using Gene Ontology and the Kyoto Encyclopedia of Genes and Genomes. A combined diagnostic model was developed using LASSO regression and validated in two independent external datasets (GSE26712 and GSE12470). Molecular docking experiments were conducted to confirm the binding affinity of PTX to key proteins. Immune infiltration analysis was performed to assess the tumor microenvironment, revealing significant differences in immune cell composition between normal and tumor tissues. A total of 2267 DEGs were identified, with 15 overlapping genes related to PTX targets. After PPI network analysis, Kaplan-Meier survival analysis, and HPA validation, five key genes (AURKA, CBX7, CCNA2, HSP90AA1, and TUBB3) were identified as associated with HGSOC progression. The combined diagnostic model demonstrated high accuracy in distinguishing HGSOC from normal tissues, with AUC values of 0.9892 and 0.9465 in the GSE26712 and GSE12470 validation datasets, respectively. Molecular docking confirmed stable binding of PTX to these key proteins, suggesting their role in PTX's therapeutic effects. Immune infiltration analysis revealed significant differences in immune cell composition between normal and tumor tissues, highlighting the potential impact of these genes on the tumor microenvironment. In summary, our findings provide a theoretical basis for improving clinical diagnosis and elucidating the underlying mechanisms of HGSOC.