Fixes and improvements for CHARMM force fields

[epretti]

• Prevent proline-specific patch from being marked as applicable to non-proline residues (added test case covering this)
• Similar fixes for glycine and proline in Drude force field (added Drude version of test case also)
• Force field should now be usable with, e.g., solvent FFXML files not specifically created with the atom type format used by convert_charmm.py: previously, improper/anisotropy assignment scripts would raise errors due to inability to parse the atom names
• Impropers/anisotropies for residues could be missed if the combination of patches that applied to a residue replaced all, not just some, of the residue's atoms: fixed this (new Drude test case covers this)
• Nicer formatting of lookup tables in improper/anisotropy scripts
• Update README to explain that custom patched version of ParmEd is needed for the conversion

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 71.41%. Comparing base (730a101) to head (34c07f0).

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #386   +/-   ##
=======================================
  Coverage   71.41%   71.41%           
=======================================
  Files           5        5           
  Lines         822      822           
=======================================
  Hits          587      587           
  Misses        235      235           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[peastman]

I looked over the changes, but I don't think I know enough to review it properly. I'll trust you on it.

Is there an open issue describing the problems this fixes?

I assume we'll want to create a 8.3.1 patch with the updated force fields?

[epretti]

Yes, I should have linked back to openmm/openmm#4976. Unfortunately the new CHARMM we released fails to parameterize systems with the existing water models that we didn't update (it raises an error instead), and the test in OpenMM that was loading the new force field doesn't use a system that has water in it, so I didn't catch it.

My plan after getting this merged into openmmforcefields was to open a PR for OpenMM that updates the force fields and adds a test that verifies that they work with the water models in OpenMM. We should probably do a patch because the CHARMM36 2024 released now only works with the water models in openmmforcefields.