Pdb2gmx for standard lipid in CHARMM; residue name causing issue?

[evanc]

GROMACS version: 2023.3 with CUDA 12.3
GROMACS modification: No

I am interested in converting a standard cholesterol .pdb file into its topology file using the CHARMM36 forcefield but am facing an issue.

I downloaded the cholesterol PDB (chl1.pdb) from the CHARMM-GUI website here.

I then attempt to run the following command where I select CHARMM for the lipid and TIP3P for water.

gmx pdb2gmx -f chl1.pdb -o chl1.gro -p chl1.top -i chl1.itp

However, I get this error…

Fatal error:
Residue ‘CHL’ not found in residue topology database

I believe this issue is emerging because the .pdb file has a 3-letter residue name in the 4th column (see first screenshot below), whereas lipid.rtp in the CHARMM36 forcefield uses the 4-letter identifier of “CHL1” (see second screenshot below)

How can I solve this issue? As can be seen in the .pdb screenshot, the rightmost column in the .pdb file has the correct residue name that matches that of the forcefield, but it seems that the 4th column with “CHL” is causing issues. I tried changing the 4th column from “CHL” to “CHL1” but this caused misalignment issues; I believe .pdb formatting only permits up to 3 characters in the 4th column.

Any tips would be appreciated! Thanks

[jalemkul]

You can add a fourth character to the residue name as long as you don’t misalign the other columns. CHL1 can be used in a PDB file.

[evanc]

Thank you. This indeed fixed the problem.

For anyone’s future reference, in the fourth column of the pdb, I edited "CHL " (note the one trailing space) to “CHL1”. (Note that this means there is no longer a space between the fourth and fifth columns which appears to cause no issues.) I initially did not include the trailing space when replacing “CHL” with “CHL1” , hence why it caused a misalignment issue.