Due to the enormous advances in computational power, quantum mechanical (QM) methods have been incorporated into the process of drug design in all its preclinical stages. Last decade, QM methods have been gaining in importance especially in structure-based drug design, where a reliable description of protein-ligand interactions is of utmost significance. However, strategies i.e. QM/MM, fragmentation or semiempirical (SQM) methods have to be pursued to overcome the unfavourable scaling of QM methods. Parametrizations of SQM and implicit solvent methods in our laboratory have been instrumental to obtain a reliable SQM-based scoring function – SQM-Score. The experience gained in its application for activity ranking of ligands binding to tens of protein targets resulted in setting up a faster scoring approach, which outperforms standard scoring methods in native pose identification for two dozen protein targets with ten thousand poses. The scoring function has also been extended to treat covalent inhibitor binding. Recently, SQM-Score was effectively applied in a proof-of-concept study in virtual screening with the aim to prioritize 72 active ligands over ~4500 inactive ones toward a HSP90 target. The SQM-Score was also tested in the first affinity ranking study that used a challenging set of 10 inhibitors binding to carbonic anhydrase II metalloprotein through Zn ion in the active site. Here, we present a large-scale affinity ranking performance of SQM-Score tested on a unique dataset of ten diverse protein systems with high-resolution crystal structures, each in a complex with ~10 ligands with known experimental inhibitory constant (K_i) or IC50. We are confident that the successful applications of SQM-Score and its feasibility and chemical generality showed its great potential to become a general tool during the hit-to-lead stage of structure-based drug design.

Adam Pecina

Istituto Italiano di Tecnologia

Appearances