Predicting drug sensitivity of cancer cells based on DNA methylation levels

Cancer cell lines, which are cell cultures developed from tumor samples, represent one of the least expensive and most studied preclinical models for drug development. Accurately predicting drug response for a given cell line based on molecular features may help to optimize drug-development pipelines and explain mechanisms behind treatment responses. In this study, we focus on DNA methylation profiles as one type of molecular feature that is known to drive tumorigenesis and modulate treatment responses. Using genome-wide, DNA methylation profiles from 987 cell lines from the Genomics of Drug Sensitivity in Cancer database, we applied machine-learning algorithms to evaluate the potential to predict cytotoxic responses for eight anti-cancer drugs. We compared the performance of five classification algorithms and four regression algorithms that use diverse methodologies, including tree-, probability-, kernel-, ensemble-, and distance-based approaches. For both types of algorithm, we artificially subsampled the data to varying degrees, aiming to understand whether training models based on relatively extreme outcomes would yield improved performance. We also performed an information-gain analysis to examine which genes were most predictive of drug responses. Finally, we used tumor data from The Cancer Genome Atlas to evaluate the feasibility of predicting clinical responses in humans based on models derived from cell lines. When using classification or regression algorithms to predict discrete or continuous responses, respectively, we consistently observed excellent predictive performance when the training and test sets both consisted of cell-line data. However, classification models derived from cell-line data failed to generalize effectively for tumors.