Comm-mode = Linear with frozen solute

User discussions
1

GROMACS version: Any version
GROMACS modification: No

Hi everyone,

as you infer from the title, I wonder whether one has some sort of bias in that way. Instead of freezing, the case of positional restraints along with comm-mode = Linear returns the following, which I have already seen in the forum:

“Removing center of mass motion in the presence of position restraints might cause artifacts. When you are using position restraints to equilibrate a macro-molecule, the artifacts are usually negligible.”

whereas freezing leads to this:

“There are X atoms that are fully frozen and part of COMM removal
group(s), removing these atoms from the COMM removal group(s)” , with X = number of atoms of your frozen moiety.

I think of two main choices:

  • comm-mode = None;
  • comm-mode = Linear and comm-grps = all_but_frozen_solute;
    which one is preferred/the correct one? Is that choice the same for both the complete freezing and the position restraints? I wonder if there is some sort of bias I am not ware of in making a choice or the other.

Thank you so much!

Bests,
Jacopo

2

I would recommend alternative two in both cases, but especially when you have a frozen group. You will then remove the center of mass motion of the part of the system that is not frozen. This is what is being done automatically according to the message above.

I don’t think there would be any drawback excluding large groups with strong position restraints from comm-removal. In principle, the center of mass motion removal might be slightly insufficient, i.e., retaining a small net center of mass motion, if you include tightly restrained atoms and they make up a significant part of the simulation.

3

Dear Dr. Magnus,

good morning, thank you again, particularly for your latter statement.

Sincerely,
Jacopo