Abstract

Drug-target binding affinity and pharmacokinetics are equally important factors of drug design. Simple molecular properties such as molecular size have been used as pharmacokinetic and/or drug-likeness filters during chemical library design and also correlated with binding affinity. In the present study, current property filters are reviewed, a collection of their optimal values is provided, and a statistical framework is introduced allowing calibration of their selectivity and sensitivity for drugs. The role of ligand efficiency indices in drug design is also described. It is concluded that the usefulness of property filters of molecular size and lipophilicity is limited as predictors of general drug-likeness. However, they demonstrate increased performance in specific cases, e.g. in central nervous system diseases, emphasizing their future importance in specific, disease-focused library design instead of general drug-likeness filtering.

Keywords: Binding site, entropy, free energy, molecule, pocket, protein, structure, target, logP, Wiener index