A comprehensive evaluation of regression-based drug responsiveness prediction models, using cell viability inhibitory concentrations (IC50 values)

• DOI: 10.1093/bioinformatics/btac177
• Publication Date: 2022-03-22T12:22:00Z
• Authors: Aron Park; Minjae Joo; Kyungdoc Kim; Won-Joon Son; GyuTae Lim; Jinhyuk Lee; Jung Ho Kim; Dae Ho Lee; Seungyoon Nam
• Keywords: Machine Learning; 0301 basic medicine; Inhibitory Concentration 50; 0303 health sciences; 03 medical and health sciences; Cell Survival; Neural Networks, Computer; Precision Medicine; 3. Good health

AbstractMotivationPredicting drug response is critical for precision medicine. Diverse methods have predicted drug responsiveness, as measured by the half-maximal drug inhibitory concentration (IC50), in cultured cells. Although IC50s are continuous, traditional prediction models have dealt mainly with binary classification of responsiveness. However, since there are few regression-based IC50 predictions, comprehensive evaluations of regression-based IC50 prediction models, including machine learning (ML) and deep learning (DL), for diverse data types and dataset sizes, have not been addressed.ResultsHere, we constructed 11 input data settings, including multi-omics settings, with varying dataset sizes, then evaluated the performance of regression-based ML and DL models to predict IC50s. DL models considered two convolutional neural network architectures: CDRScan and residual neural network (ResNet). ResNet was introduced in regression-based DL models for predicting drug response for the first time. As a result, DL models performed better than ML models in all the settings. Also, ResNet performed better than or comparable to CDRScan and ML models in all settings.Availability and implementationThe data underlying this article are available in GitHub at https://github.com/labnams/IC50evaluation.Supplementary informationSupplementary data are available at Bioinformatics online.