First report on exploring classification and regression based QSAR modelling of Plasmodium falciparum glycogen synthase kinase (PfGSK-3) inhibitors

sand quantitative prediction, respectively, of the activity of PfGSK-3 inhibitors. The classification model highlighted the contribution of electronic (Dipole-mag) and topological (S_tsC) descriptors in discriminating the PfGSK-3 inhibitors into more active and less active classes. The regression-based G-QSAR model showed the contribution of fragment-based descriptors (R1-chiV3 and R2-Most-vePotential) in determining the PfGSK-3 inhibitory activity, and also suggested modification sites for the improvement of PfGSK-3 inhibitory activity. The information obtained from this work could be utilized for the identification of novel PfGSK-3 inhibitors with a hope of overcoming the antimalarial resistance problem. © 2015 Taylor & Francis.