Machine Learning in Drug Design

Abstract

Machine learning and especially deep learning has been remarkably successful in many different areas in recent years. The application of machine learning and artificial intelligence techniques in drug discovery has increased dramatically as well. This chapter summarizes some of the most noticeable examples of the applications. Among them, the prediction of ADMET properties for drug-like molecules, de novo molecule design based on generative modeling, synthesis prediction and compound retrosynthesis planning, and in silico bioactivity predictions are the focus of this chapter. For the reader to grasp the current trend of machine learning application in drug discovery, a large number of published examples are included for in-depth analysis and discussion. Remaining challenges and potential future developments are also highlighted.