2023-01-09 Core Developers meeting notes

General updates

• DD – Should we cancel core devs on gov board days?

  • JW – I’d be happy to move it or cancel it altogether.

  • (Discussion)

  • Decision – We’ll make this every-other weekly, make it a super compact standup (aim for 15 mins)

Individual updates

• MT

  • Made 0.12.0 toolkit release, some other small features rolling in.

  • Chased down some smoke/little fires in our stack - New toolkit release caused problems in qcsubmit, some wrinkles in bespokefit,

  • Making progress on longer-term charges in interchange. Expect enough changes to have an RC in a few weeks.

• DD

  • Protein-ligand benchmarks

    • worked with Mike Henry on deployment components; not finished yet, but very close from his perspective

    • got end-to-end testing working for network submission -> transformation task generation -> task actioning -> task execution and result storage -> result retrieval

    • added user client touchpoints for compute and result retrieval

    • working this week to

      • ensure failures are handled well

      • adding GPU support to synchronous compute service

    • onboarding Hugo MacDermott-Opeskin for work on fah-alchemy

  • QCArchive

    • assembled recommendation for OpenFF QCFractal Server deployment: OpenFF QCFractal Server Deployment

• PB

  • Got back on Friday.

  • Got things together for sage paper submission to JCTC.

  • Got back into the fitting work I was doing before vacation (working with BSwope on ddE stuff)

  • CC: TG and I started using simulated annealing as a global optimizer. It showed really significant improvement. We’ve seen objective drops of around 80%.

    • CC – I’m very interested in that, let’s talk outside this meeting.

• CC

  • FB parameter fit for protein-specific model completed

  • Running QC validation for protein-null and protein-specific force fields - will give details in protein FF meeting on Thursday this week

  • Working toward setting up automated NMR benchmarks for peptides and folded proteins. I will hand this off to LWang to run on Lilac when this is ready

  • MGilson and I made substantial revisions to LiveCoMS review and sent out to external authors. Expecting comments back Monday, Jan 16.

• JW

  • Trying to clear OFFTK PR backlog, made it through about half last week

  • Updated OE license - Let me know if you have trouble with CI

  • Became convinced that it’s slightly better for fork forcebalance than not. Talking to gov board about it today.

  • Working on new PM job description in conjunction with OpenFE