Why does not dihedral angle parameters generate for new molecules?

Dear Gromacs Users,

I am doing an MD simulation of a peptide with two unnatural amino acids. I generated force field parameters using Charmm GUI and added them to chramm 36 force filed files. ( ffnonbonded.itp aminoacid.rtp and aminoacid.hbd) I also added the type of molecule to residuetypes.dat file.

I got the following errors related to improper dihedral angles when I performed gmx grompp command.

Can anyone suggest to me how to solve this problem ?

Uploading: Screenshot from 2022-02-15 21-36-16.png…

If you’re getting all these missing parameter errors, you did not add the appropriate parameters to the force field files. When you generate a system in CHARMM-GUI, it provides you everything you need. I would strongly discourage trying to take those files (which are inherently a subset of the CHARMM force field) and trying to modify the existing CHARMM port. It’s a lot of manual effort with no real purpose and likelihood of mistakes.

Thank you for your advice !
Yes, I realized it was a really hard work
May I know how to generate .itp file for the whole peptide without adding these parameters to the force filed?
I have .itp files for individual new amino acids only

If the new residues require new parameters, they have to be present in the force field. If you generated the system with CHARMM-GUI then this should already be done for you.

I understood your explanation.
Here I’m posing my peptide for simulation to get your advice.
F’PKKKKKVGHHHHL’C’HHC’HHC’GSPHHD (here F’,L’ and C’ are unnatural amino acids)
I have the .pdb file of the whole peptide and parameters for unnatural amino acids individually using Charmm GUI.
Can you propose to me how to proceed with the MD simulation in Gromacs ?

What are the modifications? The CHARMM36 force field already contains hundreds of modified amino acids - are yours not covered? If they aren’t, you have to parametrize them yourself and add any new parameters as you already attempted to, but you will have to do so very carefully. Then you need to add the modified residues to the force field .rtp file and residuetypes.dat as protein. You may also need to write .hdb entries if the coordinates you have lack necessary H atoms that need to be reconstructed. Then pdb2gmx can produce the topology of the polypeptide.

@jalemkul ,

These are the modification of my peptide sequence

I used Ligand reader and modeler to find the parameters of the above two molecules as they are not available in CHARMM 36 force field. Then I manually add these parameters to .rtp file, ffbonded.itp file and .hbd file. Still I get errors in grompp step.

My second molecule is not exactly an amino acid and I feel It cannot be modeled using CHARMM 36.
If you have any suggestions for my simulation, I would be really appreciated.

Thank you in advance !

I don’t really have any new advice here, if there are missing parameters it is because you aren’t adding everything necessary. Note that for the first compound you should only be parametrizing the side chain and leaving the Cα group alone for compatibility with the force field. You can actually assign parameters by analogy here; the sidechain is just a disulfide with an alkyl group. There shouldn’t be too many new parameters actually needed.

It’s a fatty acid with an amino group; this is very easy to model with CHARMM36. Everything is already in the force field, the only question is how it is linked within your modified peptide. I’m assuming the amino and carboxy groups mimic those of an amide in the backbone of a polypeptide?

Dear @jalemkul ,

Thank you for your guidance. Your assumption is correct for linking the second molecule with peptide chain.