GROMACS version:
GROMACS modification: Yes/No
Here post your question
Hi. Can anyone help me with the idea of applying position restrains on a system.
I have a protein and I carried out two simulations, one with restrain during NVT and other without any. However, there is no difference in the two structures I obtained after NVT. I understand usage of restrain for cases such as umbrella sampling. But for a normal simulation, should we must apply restrain ? If yes, why and how can it affect the simulation ? Can anyone help me with the understanding.
Thank you so much.
Regards
Bijaya
There are different situations when restraints are useful during equilibration. The main purpose is to avoid atoms drifting too far from their initial structure.
A common situation is that you start from a crystal (or cryo-EM) structure. If your simulation is under reasonably similar conditions and the original coordinates are good it makes sense to try to avoid large deviations from that configuration. In most cases you need to solvate this input structure, and possibly add ions. The solvent molecules (and ions) are placed without considering the macromolecular coordinates (except to avoid overlaps). In the EM stage and early equilibration stages the main purpose is to avoid atom clashes and conform to suitable bonded configurations (bonds lengths, angles and torsions) according to the force field. When using restraints the atoms in the macromolecule will be allowed to move, but they will not deviate very far from the initial configuration (the reference structure), whereas the solvent will be allowed move more (with default restraint settings) to find more favourable interaction sites, and possibly hydrogen bond networks.
This is just one example to illustrate why it can be useful. There can be other situations when you just want to prevent very large movements in early stages of equilibration, such as when starting from an “initial guess” configuration, where you might want do the equilibration in many stages with lowering restraint potentials and with the output from the previous stage as reference structure.