SMIRNOFF: CMAPs

[mattwthompson]

CMAPs are used in some biopolymer force fields but absent from the SMIRNOFF spec.

One detail that maybe should be specified is the order of the spline used to interpolate between gridpoints. There may be some differences between how engines implement this; in any case the spec should be as explicit as possible.

References:

• http://docs.openmm.org/latest/userguide/application.html#cmaptorsionforce
• http://docs.openmm.org/latest/userguide/theory.html?highlight=cmap#cmaptorsionforce
• Amber CMAP support? ParmEd/ParmEd#1066
• https://manual.gromacs.org/documentation/2019.1/reference-manual/functions/force-field.html#charmm

[SimonBoothroyd]

This would seem to be something that should be use case driven. If / when a user wants to include this, they should make their case for including it in the spec and when / if that time comes, the infrastructure team can work with the user to assess the best path forward to accommodate it.

Without a championing user, I'd recommend leaving it out. Note this does not preclude a user from creating their own CMAP handler (or anyone from making any custom handlers, it would just not be included in the spec initially.