Abstract
Materials discovery is decisive for tackling urgent challenges related toenergy, the environment, health care and many others. In chemistry,conventional methodologies for innovation usually rely on expensive andincremental strategies to optimize properties from molecular structures. On theother hand, inverse approaches map properties to structures, thus expeditingthe design of novel useful compounds. In this chapter, we examine the way inwhich current deep generative models are addressing the inverse chemicaldiscovery paradigm. We begin by revisiting early inverse design algorithms.Then, we introduce generative models for molecular systems and categorize themaccording to their architecture and molecular representation. Using thisclassification, we review the evolution and performance of important moleculargeneration schemes reported in the literature. Finally, we concludehighlighting the prospects and challenges of generative models as cutting edgetools in materials discovery.