Discovering Quality Drug Seeds by Practical NMR-based Fragment Screening

Fragment-based screening represents a potential means for smaller institutions to meet the needs for identifying the seeds for future medications. Fragment-Based Lead Discovery (FBLD) is becoming a via- ble complement and alternative to traditional high-throughput screens for discovering the seeds of future drugs. FBLD involves the screening of libraries of fragments to search for binders to target pro- teins. These binders can then be used as chemical biology probes, functional modulators or scaffolds to custom design potent inhibitors. Central to FBLD is the quality of the screening library. FBLD is a vali- dated strategy which has led to compounds in the clinic/market, but serious issues remain that limit its practical application.This work describes the practical processes employed in creating a new fragment library where a combination of theoretical cheminformatics and experimental NMR filters were employed to remove undesirable compounds (reactive, toxic, unstable, aggregators, etc.,) and to priori- tize desirable compounds (3D dimensionality, biocores, solubility, substructures).These approaches include the introduction of sensitive and simplified NMR-based screening methods, software that assists in the identification of fragment binders, and follow-up strategies that help to filter out problematic/ promiscuous ligands. Overall, screening time is reduced, deconvolution efforts are semi-automated, and medicinal chemists can focus on more promising drug seeds. Using these stringent criteria, a starting set of ~7000 compounds was reduced to an enriched subset of 1604 compounds. As an ensemble, this new library is distinct from most commercial libraries, and individual compounds are readily available along with their 1H NMR spectra in buffer. Furthermore, an intelligent pooling strategy is introduced that enables higher-throughput screening.