Oviduct Glycoprotein 1 (OVGP1) Diagnoses Polycystic Ovary Syndrome (PCOS) Based on Machine Learning Algorithms

Aims: To investigate the diagnostic value of oviduct glycoprotein 1 (OVGP1) levels for polycystic ovary syndrome (PCOS). Materials and Methods: Serum OVGP1 concentrations were measured by enzyme-linked immunosorbent assay (ELISA). Associations between OVGP1 and endocrine parameters were evaluated by Spearman’s correlation analysis. Diagnostic capacity was assessed by utilizing machine learning algorithms and receiver operating characteristic (ROC) curves. Results: OVGP1 levels were significantly decreased in PCOS patients and correlated with the serum follicle-stimulating hormone (FSH) concentration and the luteinizing hormone/follicle-stimulating hormone (LH/FSH) ratio, which are predictors of PCOS occurrence. The diagnostic value of OVGP1 combined with six signatures (LH/FSH, progesterone, total cholesterol, triglyceride, high-density lipoprotein cholesterol, and anti-Müllerian hormone) or three clinical indicators has the potential to significantly improve the accuracy of diagnosing PCOS patients. Conclusion: OVGP1 enhances the ability to diagnose when combined with clinical indicators.