Related Experiment Video
Updated: Jul 10, 2025
11:02
Identification and Quantification of Deranged Metabolites in Critically Ill Patients Using NMR-Based Metabolomics
Published on: November 29, 2024
550
Prediction of spontaneous preterm birth using supervised machine learning on metabolomic data: A case-cohort study
Yasmina Al Ghadban1, Yuheng Du2, D Stephen Charnock-Jones3,4,5
1Nuffield Department of Women's and Reproductive Health, University of Oxford, Oxford, UK.
BJOG : an International Journal of Obstetrics and Gynaecology
|November 20, 2023
Summary
Maternal serum metabolites can predict spontaneous preterm birth (sPTB). A specific lysolipid, 1-palmitoleoyl-GPE, shows promise as a novel biomarker for both sPTB and spontaneous early term birth (sETB).
Area of Science:
- Biochemistry and metabolomics
- Reproductive medicine
- Machine learning in healthcare
Background:
- Spontaneous preterm birth (sPTB) and spontaneous early term birth (sETB) are significant obstetric concerns.
- Identifying reliable biomarkers for predicting these outcomes is crucial for timely intervention.
Purpose of the Study:
- To identify and validate maternal serum metabolites that predict spontaneous preterm birth (sPTB).
- To utilize these metabolites for predicting spontaneous early term birth (sETB).
Main Methods:
- A case-cohort design was used within a prospective cohort study of 399 participants.
- Untargeted metabolomic analysis of maternal serum samples was performed at multiple gestational ages.
- Six supervised machine learning methods and Cox/logistic regression models were applied for metabolite identification and prediction.
Main Results:
- 47 metabolites, predominantly lipids, were identified as significant predictors of sPTB.
- A 4-metabolite model achieved an AUC of 0.703 at 28 weeks gestation for sPTB prediction.
- The lysolipid 1-palmitoleoyl-GPE (16:1)* emerged as a strong predictor for both sPTB and sETB across different gestational ages.
Conclusions:
- Maternal serum metabolites, particularly lipids, can be effectively identified and validated as predictors of sPTB.
- 1-palmitoleoyl-GPE (16:1)* represents a novel potential biomarker for predicting sPTB and sETB.
- External validation of these findings in diverse populations is necessary.
