Exploring acetylation-related gene markers in polycystic ovary syndrome: insights into pathogenesis and diagnostic potential using machine learning.

Objective: Polycystic ovary syndrome (PCOS) is a prevalent cause of menstrual irregularities and infertility in women, impacting quality of life. Despite advancements, current understanding of PCOS pathogenesis and treatment remains limited. This study uses machine learning-based data mining to identify acetylation-related genetic markers associated with PCOS, aiming to enhance diagnostic precision and therapeutic efficacy.

Methods: Advanced machine learning techniques were used to improve the precision of key gene identification and reveal their biological mechanisms. Validation on an independent dataset (GSE48301) confirmed their diagnostic value, assessed through ROC curves and nomograms for PCOS risk prediction. Molecular mechanisms of acetylation-related gene regulation in PCOS were further examined through clustering, immune-environmental, and gene network analyses.

Results: Our analysis identified 15 key acetylation-regulated genes differentially expressed in PCOS, including SGF29, NOL6, KLF15, and INO80D, which are relevant to PCOS pathogenesis. ROC curve analyses on training and validation datasets confirmed the model's high diagnostic accuracy. Additionally, these genes were associated with immune cell infiltration, offering insights into the inflammatory aspect of PCOS.

Conclusion: The identified acetylation gene markers offer novel insights into the molecular mechanisms underlying PCOS and hold promise for enhancing the development of precise diagnostic and therapeutic strategies.