Segmentation fault - While using ML based CHARMM force filed in gromacs

User discussions
1

GROMACS version:2023.3
GROMACS modification: Yes/No
Here post your question

I have generated force field parameters for Benzene using Machine Learning in CHARMM (CGenFF format) and would like to use these parameters in GROMACS. I have added the bonded and nonbonded parameters in ffbonded.itp and ffnonbonded.itp, respectively, defined new atom types in atomtypes.atp, and added the topology in the cgenff.rtp file (with the BZN molecule type). Using cgenff_charmm2gmx_py3_nx2.py, I was able to obtain bzn.top, bzn.itp, bzn.prm, and bzn_ini.pdb files. I also generated a .gro file from bzn_ini.pdb and solvated it. However, I encountered an issue during energy minimization:

Steepest Descents:
Tolerance (Fmax) = 1.00000e+03
Number of steps = 500000
Segmentation fault

The program shows a segmentation fault and does not proceed further.
I am attaching the relevant files, including bzn.itp, bzn.top, bzn.prm, and the solv.gro file.

bzn.itp

[solv.gro]
(solv.gro - Google Drive)

bzn.top

bzn.prm

Can someone please help me to run this system.
Thank You

2

To reproduce this we would also the mdp file.

Can you post the standard error and standard output output?

3

Dear @hess thank you for your response, Now the error has been resolved, at that time there was an issue in non bonded parameters.