To enable the establishment of a drug discovery operation for neglected diseases, out of 2.3 million commercially available compounds 222 552 compounds were selected for an in silico library, 57 438 for a diverse general screening library, and 1 697 compounds for a focused kinase set. Compiling these libraries required a robust strategy for compound selection. Rules for unwanted groups were defined and selection criteria to enrich for lead-like compounds which facilitate straightforward structure–activity relationship exploration were established. Further, a literature and patent review was undertaken to extract key recognition elements of kinase inhibitors (“core fragments”) to assemble a focused library for hit discovery for kinases. Computational and experimental characterisation of the general screening library revealed that the selected compounds 1) span a broad range of lead-like space, 2) show a high degree of structural integrity and purity, and 3) demonstrate appropriate solubility for the purposes of biochemical screening. The implications of this study for compound selection, especially in an academic environment with limited resources, are considered.

Judging if a protein is able to bind orally available molecules with high affinity, i.e. if a protein is druggable, is an important step in target assessment. In order to derive a structure-based method to predict protein druggability, a comprehensive, nonredundant data set containing crystal structures of 71 druggable and 44 less druggable proteins was compiled by literature search and data mining. This data set was subsequently used to train a structure-based druggability predictor (DrugPred) using partial least-squares projection to latent structures discriminant analysis (PLS-DA). DrugPred performed well in discriminating druggable from less druggable binding sites for both internal and external predictions. The method is robust against conformational changes in the binding site and outperforms previously published methods. The superior performance of DrugPred is likely due to the size and composition of the training set which, in contrast to most previously developed methods, only contains cavities that have evolved to bind a natural ligand.

We report on selectivity profiling of 1 in a panel of 20 protein kinases and molecular modeling indicating 1 to be highly active and selective for VEGF-R2/3. Sequence alignment analysis and detailed insights into the ATP binding pockets of targeted protein kinases from the panel result in a unique structural architecture of VEGF-R2 mainly caused by the hydrophobic pocket I, determining the molecular basis for activity and selectivity of 1.