If I want to carry out an unbound docking of the 2 chains of a hetero-dimeric protein but the individual unbound structures are not available, how do I change the structure of the bound partners to remove any native-complex structure bias that the 2 chains might have ?
I am using ZDock, a rigid-body docking algorithm. I thought of randomizing the side-chains of the surface residues of the 2 chains using GROMACS. I thought of running a simulation at a higher temperature for a short time span. This may randomize the surface side chains. The only problem is I have a large number of complexes which do not have unbound structures and such simulations take days to finish.
Are there any other methods of obtaining unbound structures from bound ones which are not time-consuming?