2024-04-03 Thompson/Wagner Check-in meeting notes

Participants

  • @Matt Thompson

  • @Jeffrey Wagner

Discussion topics

Item

Notes

Item

Notes

General updates

JW –

  • I’ve updated docs CI to high priority for JM

  • Thanks for catching the user support threads (issues and on Slack). I may engage with the Shirts lab people to understand their use case better.

    • MT – Yeah, they’re doing some tricky things going back and forth w/ GROMACS but I don’t thjink we want to have them use a different bilayer workflow using OpenMM. So I think they may have made small mistakes but mostly are on the right track.

  • I think the book is closed on the xz issue, but I also saw some grumblings about redis changing licensing for version >=8. Please forward any news/related stuff to me.

  • (MT gave some roadmap feedback, core deliverable workload seems reasonable)

MT –

  • OpenFE update - Some things are moving fast, some things are waiting on me. I just learned that OpenFF took responsibility for running validation (running compute etc). I don’t want to be in charge of nannying HPC. For the first set of validation runs, we’ve targeted 10 node-days of compute. This covers 5 different ligands in water, comparing existing and new code paths (openfe 0.15 vs 1.0). Other tests are for RG to pick 10 ligand-solvent combos and for us to run skunkworks, and then to compare to experiment. Next step would be to turn these results over to science team with the hope that we find that the new code path does similar for hydration FE, and that for other solvents it does fine too. Estimated total is 15 node-days (1 node-day per mix)

  • JW – I’ll ask “who should run these calcs” at lead team meeting later today. I think it’d make sense for science team since they’ll benefit from understanding how all this works.

  • MT – Ok, I’m happy to be an intermediary/tech support for science team doing this.

  • JW – I’d also be happy to onboard you to NRP for random compute needs and to improve bus factor.

  • MT – That’d be great, I have lots of random compute needs

  • JW – Great, please attend next week’s QC submission meeting.

MT – I’d like to take vacation the week after the symposium

  • JW – Approved

Trello

https://trello.com/b/dzvFZnv4/infrastructure

Tutorial collection

 

Action items

Decisions