International Journal of Biomedical Data Mining
Open Access
ISSN: 2090-4924
ISSN: 2090-4924
QSAR has established itself as an important means of virtual screening. Computational analysis of numerous drug molecules aided the facility to meaningfully screen specific ones for further use and experimentation. In this study, three-dimensional quantitative structure-activity relationship (3D-QSAR) associated simulations were done and stochastic models were developed considering 55 molecules of hydroxyflavanone, thaizolidene2, 4-dione, pyridone derivatives against proviral insertion site of moloney murine leukemia virus. As part of the study, molecular field analysis (MFA) was done along with receptor surface analysis (RSA). Moreover, the generalization ability of the developed model was rigorously validated using 12 test set molecules. Molecular docking exercises were performed to analyze the protein structure using the crystal structure of PIM-1 kinase. Good correlation between predicted fitness scores and the observed activities was obtained. The presence of the hydroxyl groups in B ring instead of a ring was found crucial and playing important role for the compound activity. B ring substitution showed enhanced activity in this scenario. Further, the steric descriptor coefficient present near the ortho-position of N4 amine group indicates that any steric group increases the activity. GOLD was used for molecular docking of 55 molecules considered to target PIM-1 ATP binding site. GOLD based predicted fitness scores showed excellent correlation with the observations found in the available experimental outcomes.