ABSTRACT
Background: Protein flexibility, which has been referred as a dynamic behavior has various roles in proteins' functions. Furthermore, for some developed tools in bioinformatics, such as protein-protein docking software, considering the protein flexibility, causes a higher degree of accuracy. Through undertaking the present work, we have accomplished the quantification plus analysis of the variations in the human Cyclin Dependent Kinase 2 [hCDK2] protein flexibility without affecting a significant change in its initial environment or the protein per se
Objectives: The main goal of the present research was to calculate variations in the flexibility for each residue of the hCDK2, analysis of their flexibility variations through clustering, and to investigate the functional aspects of the residues with high flexibility variations
Materials and Methods: Using Gromacs package [version 4.5.4], three independent molecular dynamics [MD] simulations of the hCDK2 protein [PDB ID: 1HCL] was accomplished with no significant changes in their initial environments, structures, or conformations, followed by Root Mean Square Fluctuations [RMSF] calculation of these MD trajectories. The amount of variations in these three curves of RMSF was calculated using two formulas
Results: More than 50% of the variation in the flexibility [the distance between the maximum and the minimum amount of the RMSF] was found at the region of Val-154. As well, there are other major flexibility fluctuations in other residues. These residues were mostly positioned in the vicinity of the functional residues. The subsequent works were done, as followed by clustering all hCDK2 residues into four groups considering the amount of their variability with respect to flexibility and their position in the RMSF curves
Conclusions: This work has introduced a new class of flexibility aspect of the proteins' residues. It could also help designing and engineering proteins, with introducing a new dynamic aspect of hCDK2, and accordingly, for the other similar globular proteins. In addition, it could provide a better computational calculation of the protein flexibility, which is, especially important in the comparative studies of the proteins' flexibility
ABSTRACT
As important therapeutic drug targets, matrix metalloproteinases [MMPs] have recently attracted great interest in the search for potent and selective inhibitors using computer-aided molecular modelling and docking techniques. Availability of more than 60 X-ray crystal structures or NMR solution structures related to MMPs in Protein Data Bank [PDB] of which more than half of them are in complex with various MMP inhibitors [MMPIs], provides a great opportunity for docking studies. In this study AutoDock 3.0.5 along with its LGA algorithm were used for automated flexible ligand docking of 32 MMPI-MMP complexes and docking accuracy and reliability of the estimated inhibition constants were evaluated. Twenty-six out of 32 docks had RMSD less than 3.0 A which is considered as well-docked, however, for the most of the cases [15 out of 27], predicted pKi values were considerably overestimated in comparison to experimental values. To improve pKi prediction regarding MMPI-MMP complexes, inclusion of at least one such a complex in calibration of empirical free energy function in the next release of AutoDock is highly recommended