2024-02-27 alchemiscale : user group meeting notes

Participants

  • @David Dotson

  • Jenke Scheen

  • Ian Kenney

  • @Mike Henry

  • @Joshua Horton

  • Meghan Osato

  • @Jeffrey Wagner

Goals

  • alchemiscale.org

    • user questions / issues / feature requests

    • compute resources status

    • current stack versions:

      • alchemiscale: 0.3.0

      • neo4j: 4.4

      • gufe: 0.9.5

      • openfe: 0.14.0

      • perses: protocol-neqcyc

      • openmmforcefields: 0.12.0

  • DD : new proposed stack versions for alchemiscale.org in time for next user group meeting?

    • alchemiscale: 0.4.0

    • neo4j: 5.16

    • gufe: 1.0

    • openfe: 1.0

    • feflow: main

    • openmmforcefields: 0.12.0

Discussion topics

  • alchemiscale.org

    • user questions / issues / feature requests

    • compute resources status

      • ~25 jobs running for OpenFE, ~1 for ASAP. Plenty of compute, feel free to submit more. We have access to MSKCC IRIS now and there’s a lot there.

      • JS – JHorton will make a big public submission for ASAP soon.

      •  

    • current stack versions (unchanged):

      • alchemiscale: 0.3.0

      • neo4j: 4.4

      • gufe: 0.9.5

      • openfe: 0.14.0

      • perses: protocol-neqcyc

      • openmmforcefields: 0.12.0

  • DD : new proposed stack versions for alchemiscale.org in time for next user group meeting?

    • alchemiscale: 0.4.0

      • not yet released, will contain some changes from IK. Will remove py2neo components, use neo4j v5. IK is also working on PRs for alchemiscale like simplifying actioning tasks, largely server-side changes. But that will be ready to be cut at any time.

      • .

    • neo4j: 5.16

    • gufe: 1.0

      • MH – not yet released, under QA currently

    • openfe: 1.0

      • MH – not yet released, under QA currently.

      • DD – No huge rush here so I’m happy to wait. Does QA include alchemiscale?

      • MH – Not yet, testing alchemiscale would be complicated for us so when we get through local QA we can talk and see who hasmore time to try this. May also depend on alchsmiscale 0.4 progress.

      • DD – Yeah, we’ll look at QA/release timing for those.

      • MH – I’ll keep you in the loop about openfe/gufe progress and code stability (like, at some point, the planned pre-release changes might just be docs)

      • MH & DD - CD would be nice to have someday

      • DD – We’ll wait until we hear from OpenFE about 1.0 RC timing. Anticipating finishing this in the next two weeks?

      • MH – Yes, I anticipate in the next two weeks, but this is assuming we don’t find any big issues.

      •  

    • feflow: main

      • Will replace perses, not super comfortable running from main so would love a release.

      • MH – Even if there’s no release, instead of pip installing from main, pick a commit hash.

        • DD – sounds good

    • openmmforcefields: 0.12.0

  • MH: openfe & gufe 1.0.0rc0 is out in the wild, QAing internally right now, when it passes internal checks I will let @David Dotson know to QA it for alchemiscale deployment

  • JW – Anyone else going to SciPy and want to prepare a tutorial application with me in the next 14 hours?

Action items

@David Dotson will replace Zoom link for this meeting and dev meeting
@David Dotson will await word from OpenFE on openfe and gufe 1.0 or RC deployment suitability before performing QA for alchemiscale.org

Decisions