MR.DRP: A Parameter Adaptive Machine Learning Framework for Drug Response Prediction

With the development of gene sequencing technology, the cost of sequencing has decreased dramatically and a large amount of sequencing data provides the basis for predicting cancer drug treatment feedback using machine learning algorithms. We address the problem of data shift between gene library data and clinical data in personalized drug response prediction tasks. Shapley Value is used to perform feature selection on differential gene features to enhance the generalization ability of the model trained. A parameter adaptive optimization mechanism using evolutionary algorithms is also integrated to propose MR.DRP, a machine learning framework for personalized drug response prediction, and validated on Cisplatin and Paclitaxel clinical data. The results show that MR.DRP achieves the best prediction results compared to other traditional machine learning models. Finally, we perform enrichment analysis at some of the important features, pointing out some of the important pathways that influence the prediction of drug response.