General updates | JM – Still working on blog post JM – Been working on COOH issue Two options JM – Best option may be to detect trans-COOHs in generate_conformers, then manually rotate them before returning the conformers, for two reasons: JW – These are good points, but reaching into generate_conformers starts putting OpenFF in the position of being a cheminformatics toolkit, which I don’t want us to be. JM will make a FrozenMolecule._make_carboxcylic_acids_cis method which will be called in OpenEyeToolkitWrapper.generate_conformers if a kwarg (make_carboxylic_acids_cis=True) is set (it’ll be False by default, but we can discuss making it True in an Issue JM will open).
JM – Also, there’s a bug in FrozenMolecule._elf_conformer_selection where the toolkit_registry kwarg fails if it’s a ToolkitWrapper. This wasn’t tested - Should I fix this in the same PR? (JM discusses a plan to have the base ToolkitWrapper expose some methods, JW doesn’t like it since it adds complexity and gives us a place where we would break external plugins by changing a base signature).
|