QSAR modeling of some anticonvulsant molecules as γ-aminobutyrate-aminotransferase inhibitors

Abstract

Abstract Quantitative structure-activity relationship study was done on 62 compounds with anticonvulsant activity in maximal electroshock-induced seizures test. The molecular structure of the compounds was optimized with parametric semi-empirical PM3 method available in Spartan 14 software. Quantum mechanical descriptors were extracted from the property and output module of the software. Combination of activity based-clustering and genetic function algorithm chemo-metric techniques were used to map molecular descriptors to activity values. A well-validated and robust quantitative structure-activity model was obtained with R 2 = 0.947, Q 2 = 0.924, F = 91.42 and R 2 pred(test) = 0.881. The descriptor contained in the model suggested an increase in the number of O and N atoms in the molecule augments the activity of the studied compounds. Also, the introduction of electron donating substituents is beneficial to the activity of the studied compounds. Armed with these, information, new hypothetical1H-pyrazole-5-carboxylic acid derivatives were designed using template approach and screened in silico . Compounds with hypothetic anticonvulsant activity better than the template were docked with γ-aminobutyrate-aminotransferase and their binding affinity was found to be comparable and even superior to that of 4-aminohex-5-enoic acid (vigabatrin), a known inhibitor of γ-aminobutyrate-aminotransferase.