How to combine param files from Charmm-gui with the CGenFF files for small molecules

GROMACS version: 2019.4
GROMACS modification: Yes/No

Dear all,

I am trying to simulate a complex consisting of lipid layers, a protein and a small molecule. I constructed the membrane-protein system in Charmm-gui and got their charmm parameter files for gromacs simulations, that is, forcefield.itp, ffnonboned.itp, ffbonded.itp etc.

Then I generated the topology against the charmm36-jul2020.ff for the small molecule by CGenFF and got the param files such as mol.itp, mol.prm. To make them work, the forcefield.itp in charmm36-jul2020.ff should be included.

The problem is the forcefield.itp from Charmm-gui and the one from CGenFF are largely different, especially the atom types. Hence I don’t know how to combine them together so that the complex can work well. Do I really need to manually select the atomtype, bondtype, angletype ect. out of many files in charmm36-jul2020.ff and then put them into Charmm-gui files?

Does anyone have a clue on this issue? Many thanks!

all the best,

Files from CHARMM-GUI are intended to be used as-is and not modified, and as such, their convention is to only include the relevant portion of the force field for the system at hand, rather than the typical approach of reading the whole force field.

Unfortunately, yes. It’s probably a better approach to simply regenerate the topology of your system natively within GROMACS using the actual charmm36-jul2020.ff files. It will be less work.

Thanks for your comments and suggestions. I regenerated the topologies for the protein and membrane within GROMACS using the charmm36-jul2020.ff. It worked very well!

Dear Justin,

I am modeling the same type of system ( lipids and small molecules ) and ran into the same problem with charmm gui and got tired of transferring ff components. So I downloaded charmm36-jul2021 and extracted ( both with linux " extract here " and tar xzf. ) Then tried to use x2top . gromacs found the ff file but called it " Charmm all atom force field." and I am certain it was the extracted file. but when selected the error arose " No or incorrect atomname2type.n2t file found (looking for ./charmm36-jul2021.ff) " but that was the extracted file. the model components came from charmm-gui. Is x2top the incorrect tool ?

We do not provide any support for x2top with the CHARMM36 port. x2top is only suitable for very simple molecules, and no aspect of converting the CHARMM force field files to GROMACS format allows for any kind of .n2t file. Don’t attempt to parametrize molecules in this way.


I see it has been over two years since this post was created. As I have been working with similar systems, I was thinking of providing my idea of how to do this. I generated the structure of a solvated membrane-protein using charmm-gui and the topology contains in total four files: forcefield.itp, protein.itp, lipid.itp, ions.itp and water.itp.

Similarly one can generate the parameters of a small molecule using charmm-gui ligand reader and modeler utility and there also you will be provided with two files: ligand.itp and another charmm36.itp (or forcefield.itp file containing the necessary atomtypes, bondtypes, pairtypes, dihedraltypes etc. only for the small molecule)

One can simply insert the small molecule into the solvated membrane-protein system by using gmx-insert and replacing water molecules. To combine the topologies, we can add two include statements in the file of the solvated membrane-protein system:

first include line–> it should include the forcefield.itp file of the small molecule (Don’t forget to comment the Defaults directive in it); this particular file should be included exactly after the forcefield.itp file of the solvated membrane-protein is defined, else gromacs will give Fatal error;

second include line–> it should include the ligand.itp file containing the ligand atom, bond etc. parameters.

In this way also one can combine the topologies of solvated membrane-protein and small molecule. Any comment or suggestion is highly appreciated.