Ligand based screening of chemical constituents from African medicinal plants for the identification of MAOB inhibitors

Abstract

Ligand based virtual screening (LBVS) is based on the hypothesis that similar structures have similar biological functions. In this research paper, ligand based virtual screening has been performed in order to predict the inhibitors for monoamine oxidase (MAO-B), an enzyme specifically involved in the metabolism of non-hydroxylated amines such as benzylamine and beta-phenylethylamine (PEA), thus, could be the target to treat various neurodegenerative disorders like Parkinson’s disease. For this purpose, Afro Database, a subset of ZINC natural compound database has been screened using Random Forest Modeling (RF). For the training of RF model, 36 reference molecules, the known inhibitors of MAO have been collected from Duke’s phyto-chemical and ethno-botanical database. As an outcome of this screening, 31 compounds out of 968 compounds from Afro Database (compounds from African medicinal plants) are predicted to be active as MAO-B inhibitor, Out of the 31 predicted active compounds, Norlichexanthone (ZINC05765089) is predicted to be most active against MAO-B with highest RF score 0.795181, along with the other top hits, could be the putative drug candidates for the prevention/ treatment of Parkinson’s disease.