Abstract
This account summarizes Dr. Subhash Basak’s work in the field of molecular similarity. In particular, it looks at the development and application of quantitative molecular similarity analysis (QMSA) techniques using physicochemical properties, topological indices, and atom pairs as descriptors for developing structure- or property-based similarity spaces and the use of a k-nearest neighbors (kNN) technique to estimate the properties or activities of chemicals within the space. Additionally, the account discusses the novel tailored similarity technique pioneered by Dr. Basak’s research group and discusses the future of molecular similarity analysis techniques.
Highlights
The concepts of molecular similarity/dissimilarity have applications in various aspects of chemistry, pharmaceutical drug design, toxicology, and ecotoxicology
This review examines the development and use of quantitative molecular similarity analysis (QMSA) methods that make use of calculated molecular descriptors derived from chemical graph theory, viz., topostructural indices (TSI), topochemical indices (TCI), and a particular class of substructures known as atoms pairs (APs)
This study resulted in the discovery that for this particular set of compounds, while the degree of overlap between the groups of analogs selected by theoretical descriptor spaces is relatively high, the similarity space constructed from physicochemical property data provided relatively unique groups of analogs with regards to those selected from the theoretically derived similarity spaces
Summary
The concepts of molecular similarity/dissimilarity have applications in various aspects of chemistry, pharmaceutical drug design, toxicology, and ecotoxicology. A perusal of the above indicates that the territory of molecular similarity is a messy area, being complicated by such factors as ambiguity in the representation of molecular structure, observation that in some cases small differences in structures may lead to large differences in biological function, and the phenomenon of bioisosterism where cellular targets can accommodate apparently widely different structures into the same site Another factor in the computation of molecular similarity is that a large number of mathematical functions can be used to derive measures of similarity for a pair of molecules starting from the same set of structural descriptors.[5,6] In defense of the usefulness of similarity, one can say that this is a widely understood intuitive notion that works well to relate molecular structure to its function both qualitatively and quantitatively in many cases. We will discuss our recently formulated idea of “tailored similarity” where the descriptor space for the computation of intermolecular similarity is property-driven instead of being intuitive or arbitrary
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
