Hi all. I have a query. It is about Whether GROMACS can do rmsd restraint of a protein throughout the whole simulation. I found Gromacs can do only position restraint. But my intention is the protein should be flexible but only restraint its conformation so that it cannot switch to a different conformation. Any suggestion will be very helpful. Thanks.
Dear Ritaban
position restraint is an option in Gromacs.
when you don't apply positions restraint, the molecules are free
when you sed this command for example:
gmx rms -f md.xtx -s md.tpr -o rmsd.xvg
the structure of a protein in trajectory will compare with reference structure in -s md.tpr file.
Hope it helps
Hi Azadeh. Thanks for your reply but my question was different. My question was ...say I have a protein in misfolded state and I want to maintain this misfolded type conformations (note that a protein can have many misfolded conformations) throughout the simulation but it should be flexible not restrained . So the way to do that is to impose a rmsd restraint on the protein ( a protein can able to explore large no of conformation with a fixed rmsd but position restraint will restrain the atoms at a fixed reference position which I don't want here) and my query was whether GROMACS can do this? Whether GROMACS can impose an rmsd restraint in protein conformation throughout the simulation?
Dear Ritaban
Thank you for your explanation.
You want to keep the misfold conformation, for this purpose the restraint should apply.
RMSD is the root mean square deviation , which evaluates deviation in position versus time,
when there is a position restraint in simulation, the RMSD does not mean for fixed molecules.
Hope it helps
Dear Dr. Halder,
This is a very interesting questions. I will give you my thoughts. You are trying to put restraints on the PDB flexibility. It is like putting the PDB in a very small box, while letting the box itself move. I believe this is kind of unrealistic in GROMACS. This is because you have forcefields for each atom. The classic definition of MD is to let the system freely move. Position restraints are like walls off of which the atoms can bounce. However, when you assign a forcefield to an atom, you are telling it to react with other nearby atoms in the same protein in a very restricted and certain way which may not agree with where the force fields take the atom. I am not sure if you are using a solution, but as usual, your system most probably includes some type of solvent. The solvent is in direct contact with your protein. Theoretically, it should be very possible to modify simulations to achieve a certain rigidity against losing a specific conformation. However, the goal of simulation is mainly seeing how the system evolves. If the protein is internally restricted, I can not see how it should be able to freely react to and basically adapt to the solvent. To put my argument in a single sentence:
"Internal restriction of PDBs, in terms of structural restrictions, may to some extent be against their free adaptation with the solvent and everything around it in that particular system, and hence the evolution of the system"
This argument may not always be true, and I may have misunderstood you. Still, I think it is perfectly possible to do what you say, but It is hard for me to see why you would put a protein in a system (probably) to see how it reacts with the system around it. And, then pervent it from even structurally getting adapted to the system. Because, in real experiments, it most certainly will not stay restricted even when forces are telling it to adapt in a certain way.
One more thing, putting RMSD restriction on your protein also seems to be kind of odd in the same way. Once the structure hits a specific RMSD value, in the next step it should just fall back to a position similar to its initial structure, even if the forces are telling it to move in a way that increases RMSD. This to me seems a little unrealistic as well, while again it is a very distinct and great question to me.
At last, I think you could put very specific position restraints on your PDB, and this would prevent it from getting away from its initial structure (this to me, again, seems against the ultimate purpose of most simulations)
Again this is a very interesting question. Please let me know if you find a way to do this, and it would be great if you could let me know the simulation scenario/question behind your question.
All the Best,
Kiarash
- Badges
- Science topic