Suppress proton warnings when building InChI map #198
Description
Using an MCL1 ligand dataset from an OpenFE demo, there are some scary-looking warnings about protons being added/removed. This probably bubbles up from how RDKit interacts with some InChI tool (a CLI?). These could be unnecessarily alarming to users who might want to use NAGL for charge assignment in other purposes, i.e. it could incorrectly signal a protonation state change when preparing a protein-ligand simulation.
Reproduction:
In [2]: from openff.toolkit.utils.nagl_wrapper import NAGLToolkitWrapper
In [3]: nagl_wrapper = NAGLToolkitWrapper()
In [4]: for molecule in Molecule.from_file("source/openfe/ligands.sdf"):
...: nagl_wrapper.assign_partial_charges(molecule, "openff-gnn-am1bcc-0.1.0-rc.3.pt")
...:
[14:59:11] WARNING: Proton(s) added/removed
[14:59:11] WARNING: Proton(s) added/removed
[14:59:12] WARNING: Proton(s) added/removed
[14:59:13] WARNING: Proton(s) added/removed
[14:59:14] WARNING: Proton(s) added/removed
[14:59:15] WARNING: Proton(s) added/removed
[14:59:16] WARNING: Proton(s) added/removed
[14:59:17] WARNING: Proton(s) added/removed
[14:59:17] WARNING: Proton(s) added/removed
[14:59:18] WARNING: Proton(s) added/removed
[14:59:19] WARNING: Proton(s) added/removed
[14:59:20] WARNING: Proton(s) added/removed
[14:59:21] WARNING: Proton(s) added/removed
[14:59:22] WARNING: Proton(s) added/removedI haven't looked into whether these are emitted directly from a Python library (easy enough to fix) or some other library to STDOUT (trickier). I think it ultimately comes from this source
Metadata
Metadata
Assignees
Labels
No labels