Protocol for generating solvated input files for various codes (AMBER, gromacs)

[jchodera]

This issue is for discussing how we should generate solvated input files for various codes.

Questions:

• Which codes should we support?
• What is our goal? To make sure the systems are as identical as possible?

[davidlmobley]

Code supported:

• My group can easily support GROMACS and AMBER
• Can your group work with us to build in support for OpenMM or similar?
• Are you aware of other codes we can easily support via maintained
  converters or anything similar? (I can check with Michael Shirts about
  DESMOND conversion from GROMACS)

Goal: To provide inputs with nominally the same parameters for several
simulation packages, with at least potential energies validated across all
supported packages.
Note: This is NOT at present a reproducibility project, though it certainly
might move in that in the future. We can cross-check initial energies for
provided systems but not the full free energies at this point.