Add support for more CHARMM force fields #1157
Conversation
swails
force-pushed
the
feature/charmm-enhancements
branch
from
f0dfe96
to
5392478
Compare
|
I am super excited to see these enhancements for LP CHARMM support. Are there any plans for a timely merge? I noticed that when I write a psf using the code in this branch for a system with lone pairs the psf section for lone pairs is missing: That leads to lone pairs without bonds to their heavy atoms. |
|
I think I'm close -- I'm basically working to make sure that OpenMM and CHARMM yield the same OpenMM system from the same files (the OpenMM code is an old version of the ParmEd code, so it's not super straightforward to translate the changes there to here). I just have to find the time to work on it. Examples help. |
swails
force-pushed
the
feature/charmm-enhancements
branch
from
463fc17
to
0762501
Compare
* Adds support for LP (lone pairs) in CHARMM PSF files * Adds a LocalCoordinatesFrame to support the CHARMM LP types * Starts to add Drude parsing to CHARMM psf parser
This is especially possible in water where a H-H bond may be added for constraints, but that may exist alongside a H-O-H angle. If both exist, the current code will add the same constraint twice which could cause problems. Now the angle constraint only adds a distance constraint if those two atoms are not *already* constrained. This fix is inspired from an OpenMM commit for the CHARMM files.
swails
force-pushed
the
feature/charmm-enhancements
branch
from
0762501
to
e944442
Compare
Properly parses the drude anisotropies with relevant tests.
These enhancements were made to an earlier version of the source code found in OpenMM, and this PR is aimed at transferring those enhancements so that ParmEd is capable of every feature supported by OpenMM.