Parametrizing a macrocyclic molecule

[shastry2358]

Hello OpenFF community! I am interested in parametrizing a macrocyclic molecule to use with the charmm36m forcefield. I have looked at tutorials in fftk to parametrize something small but nothing this complex. Would this be the right place to approach a problem like this and if so what is the best way to get started?

Thanks!

[mattwthompson]

I'm not sure by your mention of a CHARMM force field if you're trying to use that for the macrocyclic molecule or another component. If you have some code and/or partially prepared inputs we might be better able to help out.

In general, whatever you have a well-formed chemical graph - i.e. can get into an openff.toolkit.Topology - can be parametrized with a SMIRNOFF force field such as Sage 2.3.0 or a Rosemary force field (though the Rosemary line is currently in an alpha stage). Each of these use a neural net charge model which gives AM1-BCC-like charges but at approximately zero computational cost.

Molecue/topology loading/preparation tends to be the hardest part. These cookbooks are two of my favorite and might help:

• https://docs.openforcefield.org/projects/toolkit/en/stable/users/molecule_cookbook.html
• https://docs.openforcefield.org/projects/toolkit/en/stable/users/pdb_cookbook/index.html

The PTM prototype is similar to your use case (a large molecule which isn't a simple protein) and also digs into non-trivial PDB loading: https://github.com/openforcefield/ptm_prototype