Deep learning in drug discovery

Drug discovery is a process of recognizing the chemical moieties having the potential to serve as drugs. It involves the higher cost, low efficacy, and increased timelines for discovering a drug which have made it a complex process. Hence, there is an urge need of advancement in drug discovery process which can provide the revolutionary changes. In recent years, deep learning bears promise in the process of drug discovery. Deep learning plays a crucial role in various drug discovery processes namely drug monitoring, peptide synthesis, legend-based virtual screening, toxicity prediction, pharmacophore modeling, quantitative structural–activity relationship (QSAR), poly-pharmacology, drug repositioning, and physiochemical activities. This chapter presents an outline of these expanding topics related to drug discovery, the key concepts of prevalent deep learning algorithms, and motivation to investigate these techniques for their potential applications in computer-assisted drug discovery and design.