Comm-mode angular will give incorrect results when the comm group partially crosses a periodic bound

User discussions
1

GROMACS version: 2016.3
GROMACS modification: No
Here post your question
Hi.

I am trying to get the free energy profile for the interactions between two solid-slab particles by applying a pulling code (umbrella sampling).
In order to pull one solid-slab particle towards the other, I grouped all atoms in one solid-slab particle using the bond-constraint (type 6) to represent the solid particle as a molecule so that it do not cause any errors or the atoms in the particle fly away when I pulled the solid particle towards the other.

Both solid-slab particles are finite in X and Z directions but infinite in Y direction due to periodic boundary condition. I did not want to set the particle finite in Y direction in order to save computational time with a moderate simulation box.

During the pulling process, I noticed that the pulled solid-slab particle rotates around its center of mass which I do not want. So I applied the ’ comm-mode = Angular’ to remove center of mass translation and rotation around the center of mass.

The simulation ran well and produced the output files for pulling.

The only thing I am concerned is the message in the log file saying ‘comm-mode angular will give incorrect results when the comm group partially crosses a periodic bound’.

I would like to ask whether it will be a problem.

I am looking forward to your reply.

Thank you.
Best Regards,
Anh

2

Yes, I think this will be an issue.

But also I think it will be a problem to remove COM motion while you try to probe to force using the pull code which measures COM motion.

How solid should your slabs be? If they should be (nearly) solid, you can use position restraints instead and use the free-energy code to interpolate the reference coordinates between two states. dH/dlambda will then give you the force.