Mapping Thrombosis Serum Markers Using Machine Learning
Author Information
Author(s): Martins Lucas G., Manzini Bruna M., Montalvão Silmara, Honorato Millene A., Colella Marina P., Hayakawa Gabriela G. Y., de Paula Erich V., Orsi Fernanda A., Braga Erik S., Avramović Nataša, Omage Folurunsho Bright, Tasic Ljubica, Annichino-Bizzacchi Joyce M.
Primary Institution: Universidade Estadual de Campinas
Hypothesis
Can blood serum metabolites be used to distinguish between thromboembolism and antiphospholipid syndrome?
Conclusion
The study identified specific metabolites that could serve as potential biomarkers to differentiate between patients with antiphospholipid syndrome and those with venous thromboembolism.
Supporting Evidence
- Unique metabolic profiles were found between VTE and healthy controls.
- Histidine, 3-hydroxybutyrate, and threonine were identified as key metabolites.
- Machine learning models showed high accuracy in classifying patient groups.
- Alterations in metabolic pathways were linked to thromboembolism and APS.
Takeaway
Scientists looked at blood samples to find special markers that can help tell if someone has a certain blood clotting problem or not.
Methodology
The study used 1H-NMR-based metabonomics to analyze serum samples from patients with thromboembolism and antiphospholipid syndrome, applying machine learning for classification.
Limitations
The study is limited to a single-center and may not represent other populations; variability in patient profiles over time was not considered.
Participant Demographics
Participants included 32 patients each from the APS, VTE, and healthy control groups, with a majority of women in VTE and HC groups and men in the APS group.
Statistical Information
P-Value
0.4155
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website