全自动核酸提取仪，让HCD残留和支原体污染等检测更轻松&准确

Abstract

Polycystic Ovary Syndrome (PCOS) is a prevalent female reproductive and endocrine disorder. Effective and safe therapeutic strategies remain to be further explored. This study aimed to investigate the mechanisms of icariin (ICA) in PCOS. By executing weighted gene co-expression network analysis (WGCNA) across seven Gene Expression Omnibus (GEO) datasets and cross-referencing with ICA targets from five pharmacological databases, we identified 96 potential genes. An optimized machine learning framework evaluating 113 algorithm combinations and summary data-based Mendelian randomization (SMR) refined these to five candidate targets. Molecular docking and molecular dynamics simulations identified Cathepsin S (CTSS) as the primary target, demonstrating stable thermodynamic binding. In vivo, a dehydroepiandrosterone (DHEA)-induced PCOS mouse model was employed. ICA administration significantly reduced serum testosterone, restored regular estrous cycles, decreased cystic follicles, and increased corpora lutea. Concurrently, ICA treatment downregulated CTSS expression in ovarian granulosa cells. Furthermore, single-cell RNA-sequencing virtual knockout analysis predicted that CTSS ablation in distinct granulosa subpopulations might alters Wnt, NF-κB, and immune-related pathways. In conclusion, this integrated approach combining machine learning, animal experiments, and virtual knockout demonstrates that ICA improves PCOS by regulating CTSS in granulosa cells. These findings suggest that ICA is a potential therapeutic candidate for PCOS and offer new insights into its treatment.