Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Item

Notes

Merging previous PR

  • JW will share screen to walk through review process

Next Ideas on tests

  • Test data files and configuration

    • AD will post notes here on GitHub

  • AD will open 4 issues with what he noted.

  • JW – The goal of our tests is to “do the right thing” with druglike small molecules as it pertains to leveraging OpenEye and RDKit, and have a record of where there are problems. A lot of the confusion is that “druglike small molecules” and “do the right thing” are poorly defined, and previously nobody in OpenFF had enough experience to define those terms. So I’m hoping that this work can help us figure that out and make the tests more sensible.

  • AD will look into tests that load the same molecule sets using either toolkit, and where they lead to differences.

Ideas for next work

  • New test set + documenting failures

  • If we want to end up with guidance about how to prepare RDMols and OEMols to be “correctly” consumed by the OFF Toolkit, what sorts of things should we consider? How can we more safely handle these?

    • implicit Hs?

    • Aromaticity?

    • Mol fragments/radicals/dangling bonds?

    • Metadata?

    • Some background at

      Github link macro
      linkhttps://github.com/openforcefield/openff-toolkit/issues/135

...