2021-07-20 Setiadi/Wagner Check in

• JS – Tried implicit solvent parameter refitting. Fit to hydration free energies and host-guest binding. The former tends to increase the radii, the latter tends to decrease the radii (at least for N and O).

• JS – The current HCT/OBC1/OBC2 files live in test_forcefields. Could we make authoritative versions of these?

  • JW – Yes, we could put them into openff-forcefields as version 1.0.0 of their respective titles.

• JS – Would like to be able to run with GBn and GBn2 implicit solvent models.

• JS – Would also like to be able to run OBC, varying alpha/beta/gamma

  • JW – These are buried pretty deep in the OpenMM implementation. This would be kinda hard to implement, and the odds of me making a mistake are pretty high.

    • alpha/beta/gamma are baked into hard-coded OpenMM CustomGBForces for built-in implementations.

    • comments discussing alpha/beta/gamma in initial PR to add GBSA support

Prototyping new GBSA handlers

• Working in smirnoff-plugins

• JW can do most of initial implementation, but doesn’t have bandwidth/experience to validate

• We’ll do

  • Exposing alpha/beta/gamma for HCT, OBC1, OBC2

  • Adding GBn and GBn2

• JS and JW worked together to make

• JW will add GBSA force fields to the openff-forcefields package and tell JS when they’re in a release.