Quantitative clustering and matching of conformers using average electron densities

Abstract

This study presents a new approach of quantitatively clustering conformers of small molecules, such that conformers of the same group share similar electrostatic potential (ESP) maps. This helps expedite the drug design process as ESP maps guide the “key & lock” complementarity between a molecule and its receptor. The clustering is based on similarities in the average electron density (AED) of a group of interest within a molecule. AED values are computationally evaluated using the quantum theory of atoms in molecules (QTAIM). The AED tool was validated, separately, on 43 conformers of ibuprofen and 40 conformers of a tetrazole analogue of ibuprofen. The conformers were grouped based on their AED values using the K-means clustering method. It was found that conformers of the same group share similar ESP maps, with a remarkable accuracy exceeding 96%. In addition, using a drug design concept known as bioisosterism, it was found that the AED tool depicts similarities in the ESP maps of not only conformers of a single molecule, but also conformers of different molecules that share similar biological activities.