I have a question related to -r flag in gmx grompp command. From the tutorials written by Dr. Lemkul I could read that “the origin of the position restraints (the coordinates at which the restraint potential is zero) is provided via a coordinate file passed to the -r option of grompp” and it is understandable that during the NVT equilibration step we supply em.gro for -r option while running the grompp command since restraint potential is zero during energy minimization. Interestingly, while running the grompp command during NPT equilibration step (to generate NPT.tpr) why nvt.gro file is supplied for -r option as the NVT equilibration was done while applying position restraints. Importantly, when I was reading the gromacs documentation on grompp it was mentioned that “When using position restraints, a file with restraint coordinates must be supplied with -r (can be the same file as supplied for -c”. Lastly, I also looked at one tutorial regarding Molecular Dynamics Simulation of a Membrane Protein using GROMACS (A Beginner’s Guide to Perform Molecular Dynamics Simulation of a Membrane Protein using GROMACS — GROMACS tutorials https://tutorials.gromacs.org documentation) which uses the same step5_input.gro file as an input to -r every time grompp is invoked for generating tpr files for 2 NVT and 4 NPT equilibration steps.
Now I am a bit confused about which file to supply for the -r flag while running gmx grompp during NVT and NPT equilibrations specially if I am working on a system that is prepared by CHARMMGUI having multiple NVT and NPT equilibration steps. Hence I need help in understanding this.
Normally you want to supply a file with coordinates as close to the initial structure (often crystal structure) as possible. So the gro file that you used for generating the tpr for energy minimization.
Many thanks for replying. I have a followup question that while running the grompp command during NPT equilibration step (to generate NPT.tpr) why nvt.gro file is supplied for -r option in the tutorials written by Dr. Lemkul. I am a beginner and just want to make sure that I am doing the steps correctly.
Many thanks Dr. Hess for your comments and help. Just to make sure I understand this correctly, the purpose of adding a coordinate file to -r flag is to specify the coordinates for the atoms that one would want restrained before starting the simulation/equilibration so that the structure does not deviate much from the starting structure during equilibration steps. I read online and found that the position restraints require coordinate file that is used as reference to make sure that motion of the restrained atoms is restricted about those reference coordinates. Since we want the positions of atoms/coordinates as close as possible to the initial configuration/crystal structure after the equilibration steps we can supply a file to -r that has the coordinates that is as close to the initial structure. I also agree that energy minimization, NVT and NPT equilibrations will have minimal deviations from the crystal structure and hence the minimized.gro and NVT.gro can also be supplied for -r flag as I have seen in the tutorials written by Dr. lemkul.
Do let me know if I am thinking in the right direction.
Your reasoning sounds correct. In cases where you are starting from a trusted initial configuration, such as a crystal structure, you don’t want it to deviate too far from the initial structure during EM and equilibration. So, it’s a good idea to use the initial configuration as reference when applying the restraints.
However, there might also be cases where the initial structure is just a qualified guess. In those cases you might still want to restrain the structure during EM and equilibration, to avoid very large movements due to unfavourable interactions. In such a case it might be good to use the output coordinates from the previous stage during each equilibration stage, possibly lowering the restraints in each subsequent equilibration stage until you remove the restraints completely.