Metabolomic profiling and machine learning-based biomarker identification for oligoasthenozoospermia
Jinli Li1, Tangzhen Zhao2, Mengmeng Ma1
1Center for Reproductive Medicine, Shanghai First Maternity and Infant Hospital, School of Medicine, Tongji University, 2699 West Gao Ke Road, Shanghai, 201204, China.
Metabolomics : Official Journal of the Metabolomic Society
|September 17, 2025
Summary
Metabolic differences in oligoasthenozoospermia (low sperm count and motility) were identified using UPLC-Q-TOF/MS. Machine learning models accurately diagnosed this male infertility condition based on these metabolic biomarkers.
Area of Science:
- Metabolomics
- Male Reproductive Health
- Biomarker Discovery
Background:
- Oligoasthenozoospermia significantly contributes to male infertility, characterized by low sperm count and impaired motility.
- Understanding the metabolic underpinnings of this condition is crucial for developing diagnostic tools.
Purpose of the Study:
- To investigate metabolic disparities between men with oligoasthenozoospermia and healthy controls.
- To identify potential metabolic biomarkers for diagnosing oligoasthenozoospermia.
Main Methods:
- Utilized ultra-high-performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF/MS) to analyze metabolites.
- Performed differential metabolite analysis and pathway analysis (pentose phosphate pathway, TCA cycle, etc.).
- Employed machine learning models (Logistic Regression, Random Forest, SVM) for predictive analysis.
Main Results:
- Identified 211 significantly different metabolites between groups.
- Key pathways affected include pentose phosphate, TCA cycle, glycerophospholipid, and fatty acid metabolism.
- A diagnostic model achieved high sensitivity (0.93), specificity (1), and accuracy (0.97) with an AUC of 0.998 (training) and 0.963 (test).
Conclusions:
- Metabolic profiling reveals significant changes associated with oligoasthenozoospermia.
- Established a reliable diagnostic framework for differentiating oligoasthenozoospermia from controls using identified metabolites.
