Computer‐Assisted Synthesis Planning
Zhengkai Tu, Itai Levin, Connor W. Coley · 2023
Computer-aided synthesis planning (CASP) is a broad category of computational approaches designed to facilitate the selection of experimental procedures and protocols that produce a molecule of interest. CASP approaches, either data-driven or expert-driven, may operate with or without using reaction templates. This chapter introduces data-driven approaches, a terminology that has been used interchangeably with machine learning approaches (and arguably, deep learning approaches). It also introduces the wide variety of algorithms for the core CASP tasks of retrosynthesis, condition recommendation, and forward synthesis prediction. Beyond retrosynthesis, there is also plenty of scope for future work in condition recommendation and reaction outcome prediction. The chapter presents examples of different software tools and a few case studies to provide a snapshot of the capabilities of CASP tools.