2021-07-14 JW/MT Check in

Owned by Jeffrey Wagner, created
Jul 14, 2021
2 min read

Participants

  • @Matt Thompson

  • @Jeffrey Wagner

Discussion topics

Item

Notes

Item

Notes

openff-forcefields release

  • JW – This came up in the benchmarking meeting this morning, we decided to just manually cut a release of openforcefield. So I’ll close the openff-forcefields PR.

Interchange plans?

  • MT – Worked on vsites, they’re very complex. Trying to figure out how to break down the whole concept into discrete deliverables. I’ve done some experiments with sigma hole vsites, TIP4P and TIP5P waters. They’ll run in gromacs, but I don’t have them getting the right charges yet.

    • MT – Will do OpenMM and GROMACS support first, other engines will come after.

  • JW – How do we want to order GMX+OMM vsite support versus “conventional” AMBER support?

    • MT – Basically, will work simultaneously (though loosely ordered) on

      • “advanced” OMM export

      • “advanced” GMX export

      • “basic” AMBER support

        • MT – Since AMBER is largely a biomolecule package, this should go after the topology refactor

        • JW – Agree with this. Basically this is third on my big “to-do” list, after some benchmarking work and the 0.9.3 toolkit release

        • MT – AMBER formats are really complex and somewhat strange in their conventions, it seems like it’s best to continue directing people to ParmEd until we’re ready to charge into this.

  • MT – Basically having SB drive development priorities. Not sure how to gain value from Swails' involvement. I don’t think he will be willing to go elbow-deep in SMIRNOFF-land to provide feedback at this stage.

    • JW – They may be good to bring in to help design/debug interfaces. But I agree that we shouldn’t get them involved in SMIRNOFF-land.

  • MT – MS and I are planning to send monthly/quarterly update emails to Swails/Cummings/other involved parties so that they can follow along.

  • MT – No new news with VU.

    • JW – I’ve learned that the supplement isn’t eligibile for renewal, so I don’t think we should go super far to make proofs-of-concept

    • MT – Agree. I think we’re “feature complete” within the scope of our plans. So I’ll do a bit in this direction to show what CAN be done, but we don’t have direct control of grad students/users that we need to get to use this functionality.

    • JW – At this stage, we could use the new functionality to include old OPLS in openmmforcefields. Not sure if it would be useful, but we could do this.

    • MT – That seems possible, though I agree that it’s not very high-value.

    • Breadcrumbs start here:

Action items

Decisions