Having ice layer in PDB file, ERROR: Atom MW (Res ICE-1686) has mass 0 (state A) / 0 (state B)

GROMACS version: 2021.2
GROMACS modification:

Dear GROMACS users,

I am trying to simulate ice growth in presence of hBN nanosheet and ice layer in the simulation box. I use the CHARMM36m force field.
I have the coordinates for ice atoms modelled with TIP4P in the PDB file. The residue name for ice is “ICE” and I have 636 residue for ice. The ice layer is chain A. I should also mention that there will be position restraint with force constant of 1000 kJ mol−1 nm−1 on the ice molecules.
I added this residue name in residuetypes.dat file, defined this residue with its atoms and bonds in merged.rtp file as below:

[ ICE ] ; From tip4p.itp file and ffnonbonded.itp
[ atoms ]
OW OWT4 0.000 0
HW1 HWT4 0.520 0
HW2 HWT4 0.520 0
MW MWT4 -1.04 0
[ bonds ]
OW HW1
OW HW2

And I added the parameters for bond and angle in bonded parameters file in the force field as below:
[bondtypes]
OWT4 HWT4 1 0.09572000 502416.00 ; From tip4p.itp file
[angletypes]
HWT4 OWT4 HWT4 5 104.520000 628.020000 0.00000000 0.00 ; From tip4p.itp file

Also I have the below parameters in non-bonded parameters file:
#ifdef HEAVY_H
OWT4 8 9.951400 0.0 A 3.15365e-01 6.48520e-01 ; TIP4p O
HWT4 1 4.032000 0.52 A 0.0 0.0 ; TIP4p H
MWT4 0 0.000000 -1.04 A 0.0 0.0 ; TIP 4p vsite
#else
OWT4 8 15.999400 0.0 A 3.15365e-01 6.48520e-01 ; TIP4p O
HWT4 1 1.008000 0.52 A 0.0 0.0 ; TIP4p H
MWT4 0 0.000000 -1.04 A 0.0 0.0 ; TIP4p vsite
#endif

I was able to generate the topology, create the simulation box and solvate the system but in the next command line which is:
“gmx grompp -f em.mdp -c solvate.gro -r solvate.gro -p init.top -o solvate.tpr -maxwarn 1”
the system complains about the mass of the dummy atom and I got errors for each residue of ice as follows:
“ERROR 1 [file init.top, line 52]:
atom MW (Res TIP4-30) has mass 0 (state A) / 0 (state B)”

I need to generate separate posre.itp file for ice so that I can restrain ice residues.

I would appreciate if someone could help me to solve this problem.

Atoms cannot have zero mass. It’s always easiest to not change masses during during free-energy calculations.

Dear Hess,
Thank you for your respond.
Because the ice layer was modelled with TIP4P, the mass of the dummy atom is zero, and I do not know what I can do with this mass because I think it is not possible to change it. I was thinking to use the information in the tip4p.itp file which is in the force field with minor changes like changing the residue name from SOL to ICE. It did not work and it seems the system does not read this file.