Dear LAMMPS community,
To convert the force field parameters from GROMACS to LAMMPS, at several places information about bond and angle coefficients can be found out, but nothing much for dihedral.
For the Harmonic dihedral_style in LAMMPS and same dihedral kind in GROMACS, what is the relation between dihedral coefficient? Here as well division by 2 is needed like in the case of bond and angle coefficients?
Thank you
Well, the best way to make certain would be to create an input geometry where you have just a single dihedral that is not at the minimum, set all other force constants or the LJ epsilons or charges to 0 and then compute and compare energy and forces with both codes.
In addition to possible differences with respect to the definition of the force constants, you also need to consider what units are used.
Every* dihedral style listed in dihedral_style command — LAMMPS documentation clearly states its energy equation at the top of its own style documentation page. If I recall correctly, the GROMACS documentation is equally clear (albeit organised rather strangely in my view). There are quite a few subtle details to consider such as the angle direction, using cosines vs sines, and energy units, and there do exist tools out there for such conversions (although then you’d still have to verify the results for yourself anyway).
But if all else fails, simply compare the equations to determine how parameters should be converted. I’ve done this by hand spreadsheet before and it should not take more than an hour or two (plus time to validate). Do yourself a favor and document your work so that, one year from now, you don’t panic and wonder if you did it wrong.
*except none, zero, hybrid, lepton and table, for straightforward reasons.