A multi-criteria decision making approach for predicting cancer cell sensitivity to drugs

In personalized medicine for treating cancer, the most challenging task is to identify an adequate treatment for a given patient. The prediction of drug response can highly improve the clinical results and reduce the financial costs associated with chemotherapy treatment. In this paper, we present a new approach to improve the accuracy of predicting the sensitivity of cell lines to anticancer drugs. Our approach relies on selecting relevant features impacting the efficiency of predictive algorithms. We propose a multi-criteria recommendation method to identify the most accurate prediction algorithm according to the available input data. For each iteration, a different algorithm may be selected, and the optimal accuracy may be reached. The experimental results indicate that our method outperforms existing approaches in estimating the sensitivity of cell lines to different drugs.