Modified amino acid in charmm forcefield

GROMACS version: 2021.5
GROMACS modification: No


I’m trying to simulate a peptide containing a modified residue in the N-terminus, I generated the modified amino acid topology using the CGenFF webserver, and used the generated topology to modify the force field files, I created a new residue called “NPHT” in the aminoacids.rtp file, and modified the ffbonded.itp file to include the angle and dihedral parameters for my molecule. I generated the topology for my peptide using pdb2gmx - I had to use the -ter flag-, and was able to run the energy minimization and NVT equilibration, but then I noticed that my peptide separated in two, right in the modified residue petide bond. So I inspected the topology and noticed that there was no bond defined between the carbonyl carbon of my modified residue and nitrogen of the next residue, I wan’t able to figure out why this happened so I tried to just write that bond by hand, but that led to a LINCS error during the NVT equilibration, if I didn’t make that bond by hand the equilibration would run without a problem, but I need this peptide to be just one molecule, I tried to minimize minimize my start structure with a smaller step-size (0.001 and 0.0001) , and then again with the normal step-size (0.01), I tried running the equilibration with a 0.001 time-step. None of that worked.

I’m out of ideas, it has to be a problem during the topology generation step, but I don’t know what I got wrong, I’ve even done it again from scratch and got the same result.

My files are available in the following link :

Any help will be really appreciated

Thank you!

You need to add the bond to the next residue (e.g. C +N) in your [bonds] directive. You don’t have one. You can’t simply add the bond by hand in the topology because you won’t have the necessary angles, dihedrals, and pairs associated with this interaction. Update the .rtp entry and regenerate the topology.

Also note that you will have an issue with atoms O1 and O2, which pdb2gmx may interpret as chain-terminal atoms (e.g. from the C-terminus). These may get translated to O and OXT and cause problems. I suggest renaming them.

Thank you so much for your reply professor, it worked.