To develop an interpretable framework for predicting cancer drug response by integrating multi-omics data and chemical drug descriptors.
Key Findings:
Achieved an accuracy of 95.87% and an F1-score of 95.87%.
AUROC of 0.957 and AUPRC of 0.946.
Highlighted the involvement of the PI3K/AKT/mTOR pathway in drug response.
Interpretation:
The framework effectively predicts drug response while providing biologically meaningful insights, aiding in hypothesis generation and biomarker investigation.
Limitations:
The study relies on data from the GDSC2 resource, which may not encompass all cancer types.
Potential biases in the multi-omics data could affect generalizability.
Conclusion:
This study presents a novel approach to cancer drug response prediction that integrates multi-omics data with explainable AI, enhancing both prediction accuracy and interpretability.
