Updating molecular properties during early drug discovery.

Current multiparameter optimization (MPO) strategies make use of few experimental physicochemical descriptors (i.e., solubility at physiological pH and lipophilicity in the octanol/water system). Here, we show how new trends in drug discovery (i.e., large and flexible molecules for 'difficult' targets) call for the integration of ad hoc descriptors in MPO approaches. In particular, to rank, select, and optimize drug candidates, it could be relevant to have experimental data relating to the acid-base properties and the folding of the molecule to mask polar groups (so-called 'chameleonic' properties). We propose two strategies to quantify ionization and chameleonic properties and discuss their practical integration in property criteria profiles.