Metabolism-related proteins as biomarkers for predicting prognosis in polycystic ovary syndrome.

OBJECTIVE

The study aimed to explore the role of metabolism-related proteins and their correlation with clinical data in predicting the prognosis of polycystic ovary syndrome (PCOS).

METHODS

This research involves a secondary analysis of proteomic data derived from endometrial samples collected from our study group, which includes 33 PCOS patients and 7 control subjects. A comprehensive identification and analysis of 4425 proteins were conducted to screened differentially expressed proteins (DEPs). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were subsequently performed on the DEPs. To identify independent prognostic metabolism-related proteins, univariate Cox regression and LASSO regression were applied. The expression levels of these proteins were then used to develop a prognostic model, with their predictive accuracy evaluated through receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and calibration curves. Furthermore, we also investigate the correlation between clinical data and prognostic proteins.

RESULTS

The study identified 285 DEPs between the PCOS and control groups. GO enrichment analysis revealed significant involvement in metabolic processes, while KEGG pathway analysis highlighted pathways such as glycolysis/gluconeogenesis and glucagon signaling. Ten key metabolism-related proteins (ACSL5, ANPEP, CYB5R3, ENOPH1, GLS, GLUD1, LDHB, PLCD1, PYCR2, and PYCR3) were identified as significant predictors of PCOS prognosis. Patients were separated into high and low-risk groups according to the risk score. The ROC curves for predicting outcomes at 6, 28, and 37 weeks demonstrated excellent predictive performance, with AUC values of 0.98, 1.0, and 1.0, respectively. The nomogram constructed from these proteins provided a reliable tool for predicting pregnancy outcomes. DCA indicated a net benefit of the model across various risk thresholds, and the calibration curve confirmed the model's accuracy. Additionally, we also found BMI exhibited a significant negative correlation with the expression of GLS (r =-0.44, p = 0.01) and CHO showed a significant positive correlation with the expression of LDHB (r = 0.35, p = 0.04).

CONCLUSION

The identified metabolism-related proteins provide valuable insights into the prognosis of PCOS. The protein based prognostic model offers a robust and reliable tool for risk stratification and personalized management of PCOS patients.