i want to perform simulations in gromacs using gromos53a6 forcefield. when i apply positional restraints to ions. it gets invoked. but now i want to restrain bound water for which i have generated the .itp and is tying to include it in topology file. but it says "position_restraints out of bounds (1-1). this means that you have probably inserted topology section "position_restraints" in a part belonging to different molecule.....".
This section is wrong:
; Include water topology
#include "posre.itp"
; Include Position restraint file
#ifdef POSRES_WATER
#include "posre.itp"
#endif
Calling posre.itp twice in a row makes no sense, and you don't even have a water topology #included anywhere. Moreover, you won't accomplish what you want - you're restraining ALL the waters by doing this. If you only want to restrain a subset of waters, those need to be merged into the same [moleculetype] as the protein (which can be done easily with pdb2gmx) or defined separately (but that you have to do manually) and then the restraints defined based on that setup.
You can't use genrestr in this way. The only approach that will work is the one I have said in my reply above.
This can (technically) be done, but you need to stop and ask yourself whether such a setup is valid. If the simple diffusion of water molecules is causing structurally bound Ca2+ ions to dissociate, then either (1) the force field is inadequate or (2) the experimental assignment of Ca2+ positions is wrong, at least in the context of a fully solvated simulation. In either case, you're just putting a Band-Aid on a symptom to fit some preconceived notion, and an absolute position restraint on the motion of waters will likely give rise to artifacts in the dynamics of nearby side chains.
There is considerable literature on divalent ion simulations; you should spend more time evaluating what force field(s) are sufficient, if any. Then evaluate the experimental information available to you to decide if what you're seeing might actually be legitimate.
I described how to do that in my first post. The restraints are still a bad idea. If the sites are of insufficient strength to preserve the interaction, an absolute restraint will hold the Ca2+ in place, but this time, the protein would simply drift away. You should really investigate different force fields, and also realize that experimentally assigned ion positions can often be wrong; sometimes they're just water that get incorrectly assigned based on assumptions or poor resolution.
thanks for the suggestion sir....will definitely take ur point into account and try accordingly
- Badges
- Science topic