Hii
I performed a Protein-peptide docking in HDOCK server whose binding sites are not known yet, the pdb obtained as result (docked one having highest score) was simulated for all atom MD for 1000ns using OPLS force field.After 1000ns I saw some of the residues are in bound form throughout the trajectories but these are not same as found from HDOCK.
Then I took the pdb of 1000ns and perform all atom MD for 1000ns using CHARMM forcefield and found the binding residues again different and also they are not in bound form throughout the trajectories i.e they are highly fluctuating.
So my question is OPLS forcefield is relaible for such systems or not ? or any suggestions for this will be highly appreciable.
Running the same simulation twice also may lead to different result, so whether the difference is accidental should be considered first.
Different forcefiled need to be used with different condition, such as the water model, rvdw, rcoulomb and other parameters. For example, CHARMM prefer to TIP3P and OPLS prefer to TIP4P. So make sure that every parameter is suitabel in the two simulation.
As I know, OPLS is not usually used for protein-water systems, because Amber14sb+TIP3P or CHARMM+CHARMM-modified TIP3P is a better choice. But if you have enough reason to use OPLS, you can ignore this.
I hope these could help you.
I don’t think that the difference is caused by the water model.
I think it is either because different simulations can produce very different results and/or because different force fields can favor different states.
Dear hess
I think if the obtained result is the binding sites for that particular peptide in OPLS it should show similar result even we are using different force field as both simulation is done for the same docked system.Is there possibility of capturing local minima in case of OPLS forcefield ?