Removing the all intramolecular interaction for solute molecule Or just freezing solute

User discussions
1

GROMACS version:
GROMACS modification: No
How do I remove all forces (bond stretch, bending, dihedral, LJ occurs internally etc.) for one solute molecule while keeping the LJ interaction on for solute and solvent molecule?
And related question What does following line (in [moleculatype] section) does to simulation? My naive understanding is following line will eliminate all non bonded interaction to make LJ and rescaling it to zero interaction. However, I do not think this will eliminate physically connected bonds and angles including dihedral (less than 3 atoms away). Yes for listed 1-4 interactions in topol.top only?
; nbfunc comb-rule gen-pairs fudgeLJ fudgeQQ
1 1 yes 0.0 0.0
This line still gives me some energy to isolated molecule so I think I am wrong in my interpretation.

Is this remove the rest of the intra-molecular forces?
;NAME nrexcel
Protein_Chain_A. 5

Alternatively, I saw the Freeze group option in The Group Concept (p15 in manual). I do not understand how to use this option. Any example would be greatly appreciated. Can I use this for the entire solute molecule? If so, how do I implemented?
I tried to add these line to M.mdp file
freezegrps = Protein
freezedim = Y Y Y
for this option, it crashes at first line at terminal
gmx_d grompp -f M.mdp -c G.gro -r G.gro -p topol.top -n G.ndx -o em.tpr
gmx_d mdrun -v -deffnm em

Thank you very much.

2

You do not write what you want to achieve and why.
If I understood you correctly, you want to remove all intra-molecular interactions in the solute, but keep all interactions between solute and solvent. Why do you want to that?

3

Thank you very much for the reply. I am trying to freeze solute during the simulation and my solvation free energy value (small energy in terms of magnitude) has large error when I use position restriction value of 100000.(my attempt to keep it freeze). I thought there may be a potential indirect complications when I use large force constants because atoms in the solute still move and that may result in large energy fluctuations when I repeat the simulations many times. So I thought I maybe able to eliminate (or potentially identify) the issue in my simulation since internal energy has nothing to do with free energy perturbation if there is a better way to freeze the solute.

4

Removing intramolecular forces will not freeze the solute.

Note that restraining or freezing the solute will affect the solvation free energy. I think a restraint force constant of 100000 will cause instabilities. You could try 10000. Freezing the solute should also work. You will need to set ref-coord-scaling=com when using pressure coupling.

5

Thank you very much. I will use 10K instead. And yes I am using center of mass for pressure coupling for pressure coupling.