AI and Machine Learning in Drug Discovery
Harry C. Yang, Feng Dai, Richard Baumgartner · 2022
This chapter concerns the progress of those artificial intelligence (AI) applications and the outlook of AI in drug discovery. Drug discovery is the first and crucial step of the value chain of drug development and involves target identification, optimization, and validation through preclinical testing through cell-based assays and animal models. Traditionally, compounds extracted from natural sources have played a central role in drug discovery. Like the small-molecule drugs, biologic drug discovery commences with the understanding of the disease, screening of a large number of compounds, and using both the traditional and rational approaches. By far, applications of AI in drug discovery have been largely focused on machine learning (ML) and deep learning (DL). ML and DL have been used in numerous cases for target identification and validation; compound property and activity prediction; de novo design; prediction of drug–target interactions; chemical synthesis planning; and computational pathology.