AbstractBioisosterism is a key concept in medicinal chemistry, as it allows medicinal chemists to interchange structural fragments without significant perturbation in biological activity. Not surprisingly, given the vast amount of bioactivity data and chemoinformatics resources now available, there has been a significant surge in the number of computational approaches available to mine and identify bioisosteric replacements for fragments of bioactive compounds. Such methods have certainly provided medicinal chemists with a diverse arsenal of in‐house, commercial, and academic tools and interfaces to aid in the optimization across a number of end points such as bioactivity; selectivity; and absorption, distribution, metabolism, excretion, and toxicity properties for effective and efficient drug design. These in silico bioisosteric replacement mining approaches can generally be divided into two categories, namely ligand based and structure based. The approaches of the former category use information that is derived from ligands, whereas the latter category requires knowledge of the biological target, as well as specific knowledge of the interactions between ligand and target in the binding pocket. Ligand‐based methodologies are also typically divided into similarity‐based and database mining (or knowledge‐based) approaches. In general, the former provide an assessment of putative bioisosteric fragments and substructures, in terms of molecular topology and descriptors, whereas the latter extract structural transformations from large chemical repositories and associate them with the induced change in biological or any other property of interest. Following systematic retrospective studies, a large number of nonclassical bioisosteric equivalents have been reported in the literature.This article is categorized under: Computer and Information Science > Chemoinformatics