QSAR modeling of dihydrofolate reductase inhibitors as a therapeutic target for multiresistant bacteria

Abstract

Antibacterial resistance is a growing public health threat of major concern around the world so development of new therapeutic approaches to prevent bacterial multidrug resistance has become a primary consideration for medicinal chemistry research. QSAR models for the dihydrofolate reductase inhibition with 2,4-diamino-5-(substituted-benzyle)-pyramidine derivatives were developed with further computer-aided design of new derivatives with desired activity. The Monte Carlo method has been used as a computational tool for QSAR modeling. For the representation of molecular structure and optimal descriptor calculation, the simplified molecular input line entry system (SMILES) together with the molecular graph (hydrogen-suppressed graph-HSG, hydrogen-filled graph-HFG, and the graph of atomic orbitals-GAO) was used. One-variable models have been calculated for one data split into training, test, and validation set. The impact of Morgan's extended connectivity index on built QSAR models and outliers was determined. Statistical parameters for the best QSAR model are satisfying. Structural indicators (molecular fragments) responsible for the increase and the decrease of the stated activity are defined, and with the application of defined structural alerts, the computer-aided design of new derivatives with desired activity is presented. Computational experiments presented and applied in this research can satisfactorily predict desired endpoint and can be used further for computer-aided antibacterial drug design.