I don't think I can perform Monte Carlo in explicit water.
Don't you have to re-equilibrate the water molecules around the protein after each move? Or am I wrong? I would be happy for any advice. Thanks.
In an explicit water MC simulation, the water molecules should also be subjected to Monte Carlo moves. Under these circumstances, the original Metropolis Monte Carlo method is expressly designed to yield a canonical ensemble, and variants for other ensembles are available in CHARMM. Over the last 50 years, MC has extensively been used for studies of bulk liquids and solute-solvent systems.
The catch is that the magnitude of the moves must be sufficiently small to get a reasonable acceptance rate; I don't know how feasible it is to achieve your goal with such small moves. It would be like doing an explicit water MD simulation, only faster (but not necessarily fast enough for your purpose).Edit:
I have another problem using the Monte Carlo module in CHARMM. There must be errors while computing the energy because the Delta-E column in the energy output contains values between 0 and several thousands. Delta-E is the difference between the MC running total and the current total (according to mc.doc). I have set "INBFrq -1 IMGFrq -1", but I have also played with different values between 0 and 100.
The errors are bigger with larger Monte Carlo moves.
Your setup is unusual to me, but wouldn't all of that that be expected behavior?
At the beginning of the Monte Carlo run, the protein spins around fast and the values in Delta-E are large, but once the protein is in contact with the surface the values in Delta-E are smaller and become 0 once the protein stops moving.
Unless you're annealing (which I don't see in your input), this is not supposed to happen. I speculate you're ending up in a local minimum and your move size is so large that every step is rejected by the Metropolis criterion. This is typically not desirable because it prevents you from sampling other minima. If so, you probably want to increase the MC temperature in order to cross some barriers; as you're keeping your protein frozen and you don't seem to be aiming to recreate any kind of physical ensemble, the MC temperature can be thought of as a parameter that determines the acceptance rate.
I have played around with all possible parameters but can't figure it out.
Trying to gain insight into the problem by reading might in this case be more effective than a manual MC search of parameter space...