Aim and Objective: Vitamin D3 (1,25(OH)2D3) is a biologically active metabolite and
plays a wide variety of regulatory functions in human systems. Currently, several Vitamin D
analogues have been synthesized and tested against VDR (Vitamin D Receptor). Electrostatic
potential methods are greatly influence the structure-based drug discovery. In this study, ab inito
(DFT, HF, LMP2) and semi-empirical (RM1, AM1, PM3, MNDO, MNDO/d) charges were
examined on the basis of their concert in predicting the docking pose using Induced Fit Docking
(IFD) and binding free energy calculations against the VDR.
Materials and Methods: Initially, we applied ab initio and semi-empirical charges to the 38 vitamin
D analogues. Further, the charged analogues have been docked in the VDR active site. We generated
the structure-based 3D-QSAR from the docked conformation of vitamin D analogues. On the other
hand, we performed pharmacophore-based 3D-QSAR.
Results: The result shows that, AM1 is the good charge model for our study and AM1 charge based
QSAR produced more accurate ligand poses. Furthermore, the hydroxyl group in the side chain of
vitamin D analogues played an important role in the VDR antagonistic activity.
Conclusion: Overall, we found that charge-based optimizations of ligands were out performed than
the pharmacophore based QSAR model.