An Artificial Intelligence Approach Based on Hybrid CNN-XGB Model to Achieve High Prediction Accuracy through Feature Extraction, Classification and Regression for Enhancing Drug Discovery in Biomedicine

Abstract

In the field of biomedicine, drug discovery is the cycle by which new and upcoming medicines are tested and invented to cure ailments. Drug discovery and improvement is an extensive, complex, and exorbitant cycle, settled in with a serious extent of vulnerability that a drug will really be successful or not. Developing new drugs have several challenges to enrich the current field of biomedicine. Among these ultimatums, predicting the reaction of the cell line to the injected or consumed drug is a significant point and this can minimize the cost of drug discovery in sophisticated fashion with a stress on the minimum computational time. Herein, the paper proposes a deep neural network structure as the Convolutional Neural Network (CNN) to detain the gene expression features of the cell line and then use the resulting abstract features as the input data of the XGBoost for drug response prediction. Dataset constituting previously identified molecular features of cancers associated to anti-cancer drugs are used for comparison with existing methods and proposed Hybrid CNNXGB model. The results evidently depicted that the predicted model can attain considerable enhanced performance in the prediction accuracy of drug efficiency.