In 2012, high-resolution structures were available for about 800 ligand binding pockets in human proteins, frequently in multiple conformations. As presented on page 1669 by Andrey V. Ilatovskiy et al., the background bubble plot schematically represents each pocket in the human structural pocketome as a sphere whose size is proportional to the size of the corresponding conformational ensemble. This rapidly growing data set creates remarkable opportunities for computational prediction of novel protein-ligand interactions in the drug research applications outlined in the foreground image. While ab initio quantum mechanical approaches are known to provide unprecedented accuracy in structure-based binding energy calculations, they are limited to only small systems of dozens of atoms. In the structural chemogenomics era, it is critical that new approaches are developed that enable application of QM methodologies to non-covalent interactions in systems as large as protein-ligand complexes and conformational ensembles.