Drug discovery and potential gene and pathway associated with polycystic ovary syndrome through text mining and biomedical databases

Objectives: Polycystic ovary syndrome (PCOS) is a common endocrine disease in women of childbearing age. Although it is a leading cause of menstrual disorders, infertility, obesity, and other diseases, its molecular mechanism remains unclear. This study aimed to analyze the target genes, pathways, and potential drugs for PCOS through text mining. Methods: First, three different keywords (“polycystic ovary syndrome”, “obesity/adiposis”, and “anovulation”) were uploaded to GenCLiP3 to obtain three different gene sets. We then chose the common genes among these gene sets. Second, we performed gene ontology and signal pathway enrichment analyses of these common genes, followed by protein–protein interaction (PPI) network analysis. Third, the most significant gene module clustered in the protein–protein network was selected to identify potential drugs for PCOS via gene–drug analysis. Results: A total of 4291 genes related to three different keywords were obtained through text mining, 72 common genes were filtered among the three gene sets, and 69 genes participated in PPI network construction, of which 23 genes were clustered in the gene modules. Finally, six of the 23 genes were targeted by 30 existing drugs. Conclusions: The discovery of the six genes (CYP19A1, ESR1, IGF1R, PGR, PTGS2, and VEGFA) and 30 targeted drugs, which are associated with ovarian steroidogenesis (P <0.001), may be used in potential therapeutic strategies for PCOS.