[gmx-users] Impropers in OPLS-AA

[Ignacio Fernández Galván]

Hi all,

I'm trying to build a topology file for a new molecule (benzimidazole) with the OPLS-AA force field, and I'm confused about which dihedrals and impropers I should include. I'm assuming that the molecule is similar enough to indole and adenine/guanine to use their parameters (except maybe their atomic charges, but I'll disregard this for the moment).

I've decided to build an .rtp file and use pdb2gmx to create the topology. All is fine with bonds, angles and dihedrals, I think. In the .rtp I include the atomtypes and bonds, and all connected angles and dihedrals are automatically generated, and I guess that's the right thing here. But which impropers should I include? (I mean, what is the default in OPLS-AA, or what is it "designed" for). I've tried to find it in the OPLS literature, but couldn't find a clear indication of which exact terms to include in a (arbitrary) molecule.

I tried to look at the TRP residue in the aminoacids.rtp file, but it only raised some doubts:

1. There are no defined impropers around the bridge atoms CD2 and CE2.

2. Each improper has the central atom as the 3rd atom, and not the 2nd, as the macro names suggest (improper_Z_CA_X_Y). Maybe this does not matter?

So, can anyone tell me (or point me to somewhere where I can read it) whether I should include *all* impropers around aromatic atoms and which is the right order for OPLS-AA?

Thanks,
Ignacio