Biomarker discovery and development of prognostic prediction model using metabolomic panel in breast cancer patients: a hybrid methodology integrating machine learning and explainable artificial intelligence
2024

Breast Cancer Biomarker Discovery Using Metabolomics and Machine Learning

Sample size: 201 publication Evidence: moderate

Author Information

Author(s): Yagin Fatma Hilal, Gormez Yasin, Al-Hashem Fahaid, Ahmad Irshad, Ahmad Fuzail, Ardigò Luca Paolo

Primary Institution: Inonu University, Türkiye

Hypothesis

This study aims to identify targeted metabolomic biomarker candidates for the specific detection of breast cancer using explainable artificial intelligence.

Conclusion

The study identifies potential biomarkers for early breast cancer diagnosis and demonstrates the effectiveness of targeted metabolomics combined with machine learning.

Supporting Evidence

  • Machine learning models showed high precision, recall, and specificity for breast cancer classification.
  • SHAP values identified key metabolites like leucine and isoleucine as important biomarkers.
  • The study utilized a hybrid methodology combining machine learning and explainable AI for better interpretability.

Takeaway

Researchers found specific substances in the blood that can help detect breast cancer early, which could lead to better treatment options.

Methodology

The study used plasma samples from breast cancer patients and healthy controls, applying machine learning models and SHAP for feature selection.

Limitations

The study lacks an independent validation cohort, which limits the generalizability of the predictive models.

Participant Demographics

102 breast cancer patients and 99 healthy controls, matched by age.

Statistical Information

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.3389/fmolb.2024.1426964

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication