A key question in approaching a new target is whether it is possible to develop a small molecule drug that can bind. This is frequently referred to as the “druggability” of the target. Acpharis and Boston University today published an article in the Journal of Medicinal Chemistry that presents a computational method for analyzing the druggability of a target and expands the definition into a new class of targets that are “non-canonically druggable” by new classes of molecules such as macrocycles and folders.
It has long been known that the hit rate of fragments in an NMR screen is a predictor of the druggability of a target. Analogously, we have shown in previous work that the hit rate of fragments is able to predict the druggability of protein-protein interactions. A hot spot from mapping with a large number of probes is necessary for a protein to be druggable. In our work published today, we expand this by examining the geometry of the hot spots in the binding site to show sites are either too large or too small for canonical drug-like molecules. In these cases, one must expand the library used to include molecules that do not fit the rule of five.