Issue with pdb2gmx and charm36mut forcefield: Undefined Atom Type "OH"

Hello everyone,

I am encountering an issue when using pdb2gmx with the charm36mut forcefield. The error message I receive is: “Atom type OH specified in terminal database has not been defined in the force field.” I have checked the forcefield.atp file and found an entry for “OH1”. I tried renaming the “OH” atoms in my PDB file to “OH1”, but the issue persists.

Interestingly, the error still occurs even when there are no atoms named “OH” or “OH1” in the PDB file. Furthermore, I experience the same problem when using the tripeptide.pdb files for the charmm36 forcefield, which I downloaded directly from the official website.

Has anyone encountered a similar issue or have any suggestions on how to resolve this problem?

Best regards,

Hi Marc,

– “I tried renaming the “OH” atoms in my PDB file to “OH1”, but the issue persists.”
: “OH1” is an atom type, so renaming at atom name to “OH1” will not help.

– “Interestingly, the error still occurs even when there are no atoms named “OH” or “OH1” in the PDB file.”
: atom type “OH” is usually assigned to hydroxyl oxygen in a standard force field (not mutation ff). It doesn’t matter if atoms are named “OH”/“OH1” or not.

– “have any suggestions on how to resolve this problem?”
: It seems like there is some mix-up of force-field parameters during topology generation. Could you check if the error is reproducible in a fresh/new terminal and start the protocol again from scratch? If you still see the error, please paste the steps you followed and the errors here for better understanding. (Don’t forget to “export GMXLIB=path_for_pmx_installation/pmx/src/pmx/data/mutff/”)


Dear Sudarshan,

I appreciate your response and apologize for my delayed reply. Regrettably, I am still encountering the error even after trying a new terminal. Reproducing the error is quite simple:

gmx pdb2gmx -f ${tripeptide}.pdb -o hybrid_conf.pdb -ff charmm36mut -water tip3p

Here, ${tripeptide} represents any tripeptide.pdb file downloaded from the website for the CHARMM36 force field. Although I understand that tripeptide topologies can be obtained from the website, the error persists for my actual protein data, which I convert to GROMACS format using:

gmx pdb2gmx -f ${protein}.pdb -ignh yes -o ${protein}.gro

Subsequently, I perform mutation using the PMX script: -f ${protein}.gro -o mutant.pdb -ff charmm36mut

For the time being, I have switched to the Amber99sb-star-ildn-mut force field, which functions as expected.

Thank you for your time and assistance.


To the pmx ‘develop’ branch I pushed several scripts: and

These might help you prepare the structures for a specific force field

1 Like

Awesome, thank you!