RNNDRP: Developing Recurrent Neural Network Based Anticancer Drug Response Prediction Model

Due to the devastating nature of the cancer disease, patients have a high death rate in the worlde. For this purpose, pharmacologists, medical scientists, and biologists are making collaborating efforts to efficient medications for the treatment of cancer. In the past, all cancer types were treated with the same approach in a one-size-fits-all manner. In this research, we present Recurrent Neural Network-based Prediction Model (RNNDRP) that accurately predicts drug sensitivity on cell lines data of cancerous patients using GDSC dataset for training and testing. Our prediction model accurately predicts minimal inhibitory concentration (IC<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">50</inf>) for various drugs. Our RNNDRP model has obtained the RMSE and R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> scores of 0.814 and 0.875, respectively. The performance has shown to be better than existing state of the art models developed for drug sensitivity. These improved results validated the model's potency due to hybrid features in addressing the problem of anticancer drug response prediction. The proposed model results might reduce the cost and time elapsed on lab tests for cancer treatment.