Make REST2 scale AMAP potential

[jlmaccal]

The REST2 transformer should scale the AMAP potentials along with the dihedrals.

[jlmaccal]

This should be straightforward.

We need to:

• Find the AMAP force in `finalize'
• Save the original maps in finalize
• Update the force with scaled versions of the maps in update

[cing]

It's probably worth noting that the AMAP potentials packaged with MELD were only parameterized for use with implicit solvent. Specifically, ff12sb+gbneck2 (see: https://www.ncbi.nlm.nih.gov/pubmed/26574266). So, enabling support for REST2 with AMAP might not be a good idea in this specific case but there's no harm in supporting it. As a side note, even ff14sbside has never been properly validated with these AMAPs.

[jlmaccal]

Actually, this is a good point. We probably shouldn't even allow use_amap to be set with explicit solvent.

[cing]

Scaling CMAP for REST2 still seems like it should be supported in preparation for eventually having CHARMM with CHARMM CMAPs working (which were parameterized with explicit solvent).

[jlmaccal]

OK, let's do both then. I'll add a PR for the compatibility check.

[alberto99]

The new generation of Amber post ff14 will also have cmaps for explicit solvent.

…

On Sun, Jul 15, 2018, 01:52 Justin MacCallum ***@***.***> wrote: OK, let's do both then. I'll add a PR for the compatibility check. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#44 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAM1nOPOTSY8MSNnQvGzH6nfUWxP2tuyks5uGtisgaJpZM4VPtvF> .

[jlmaccal]

The new generation of Amber post ff14 will also have cmaps for explicit
solvent.

Right, but those will be added directly by Amber. So we'll have to scale them, but we won't have add them ourselves. This will probably required updating the Amber parser in OpenMM, as I don't think Amber currently supports CMAPs in Amber format files?