Improving Drug Response Predictions in Breast Cancer with RNA Editing
Author Information
Author(s): Bernal Yanara A., Blanco Alejandro, Oróstica Karen, Delgado Iris, Armisén Ricardo
Primary Institution: Universidad del Desarrollo
Hypothesis
Can integrating RNA editing data with multiomics improve machine learning models for predicting drug responses in breast cancer patients?
Conclusion
Incorporating RNA editing into predictive models enhances the accuracy of drug response predictions for breast cancer patients.
Supporting Evidence
- RNA editing data improved the performance of the random forest model for predicting drug response.
- The best-performing model achieved an F1 score of 0.96.
- Three RNA-edited sites were identified as significant predictors of therapy response.
- The risk score for nonresponse to therapy was significantly lower in responders compared to nonresponders.
Takeaway
This study shows that looking at RNA editing can help doctors better predict how breast cancer patients will respond to treatments.
Methodology
The study analyzed data from 104 breast cancer patients, using machine learning models to predict drug responses based on clinical and omics data.
Potential Biases
Potential biases due to small sample sizes and methodological issues in previous studies.
Limitations
The study's findings may not be generalizable due to differences in preanalytic processes and RNA sequencing methodologies.
Participant Demographics
Patients were characterized based on therapy response, molecular subtype, tumor size, nodal status, histological type, and age group.
Statistical Information
P-Value
<0.001
Confidence Interval
95% CI: 0.957–0.961
Statistical Significance
p<0.001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website