How does docking in AutoDock and AutoDock Vina influence the protein itself, especially binding site? Does it change at all? I'm asking because if the protein is affected, I'd like to be able to compare it with original conformation, before docking.
Unless you define the flexible residues of the receptor, the protein structure will be kept rigid (it does not change after docking procedure) while the ligand is often kept flexible (torsions are defined while preparing PDBQT files).
Vitor Rabelo, is there a way in such case to compare the residues selected as flexible before and after docking? Is there any structure file generated by AD / ADV?
Yes, if you choose flexible residues in the protein in the output file for each docked pose you will have the coordinates of both the ligand and the flexible residues. You can then use any standard tool (e.g. those from pymol) to calculate the rmsd versus the original structure.
Obviously the differences will be only in side-chains of those residues.
I would echo the reply from Bartosz. One thing that you might want to look at while docking is the force field applied by the program. Some docking programs might not include non-polar hydrogens that gives an often incorrect estimation of the conformation. If you choose to perform MD or QM/MM calculations on the docked structure, then that problem might be solved because of treatment at higher levels of theory than normal scoring.