Generation of .itp files for small molecules for OPLS and AMBER force fields in the GROMACS package

Dear users
I want to simulate a protein-ligand system containing a small molecule in the GROMACS package using both the OPLS and AMBER force fields. For this, I need to generate .itp files for the small molecule compatible with each force field. I have tried generating them using LigParGen for OPLS and acpype for AMBER, as suggested by GROMACS tutorial.

However, the generated files contain atom types like opls_818, opls_839, etc., for OPLS, and p5, oh, etc., for AMBER, which do not appear to be directly compatible with GROMACS.

Could anyone suggest an alternative approach or point out if I am doing something wrong? Any help would be greatly appreciated.


Your issue is not that OPLS and AMBER aren’t directly compatible with GROMACS (which they are), but that you’re mixing the two, which is generally a bad idea. You have a bunch of OPLS atomtypes in your [ atomtypes ] directive, and then you’re adding AMBER-style atomtypes in the [ atoms ] directive.
E.g., if you’re using OPLS, the type part of your atom definition should be something like opls_XXX. Both LigParGen and acpype should generate .top and .itp files that are intended to be used together, so just deciding for one of them should do the trick.

Hi lmullender

Thank you for your response.

Actually, I was generating the .itp file separately for the molecule using two different forcefields for two different MD simulations. I first tried the molecule in LigParGen for OPLS forcefield, which generated atom types like opls_818, opls_839, etc. However, I also need the .itp for the Amber forcefield for another simulation, which generated atom types like p5, oh, etc. Unfortunately, both of these are not compatible with GROMACS 2020.4.

Hi Devid,

Does that mean you tried it with other GROMACS versions and it worked? I don’t know what exactly your steps are, but like I said, the issue is not that OPLS and Amber force fields aren’t compatible with GROMACS, but that they not compatible with each other. That means, if you’re using e.g. LigParGen/OPLS to generate your ligand.itp file, you should also use the OPLS force field for the rest of your topology. GROMACS doesn’t care what the atom types are called, as long as they match. If you only tell it about the OPLS atom types like opls_XXX, but then you add atoms that have an Amber-style atom type (e.g. p5, oh), gromacs won’t know what to do with that.

If you’re willing to share your .top and .itp files, maybe we could help you in more detail.