2023-03-28 Protein-ligand benchmarks meeting notes

IA : protein-ligand-benchmark : establish new working group, or use this one?

• IA – potential new participants -

  • Mirai Kreia (?) from OE is interested in helping with and running PL benchmarks

  • Meghan Osato from Mobley lab will be running stuff

  • Iván had some systems he wanted to add

• IA – Should that be a separate working group/in this same meeting?

  • DD –

  • JW – Should also answer “who’s in charge?” - Could assign a single individual, multiple individuals ANY of whom can approve, or multiple individuals ALL of whom must approve.

  • IP – We have an open issue with Toni Mey about governance of this dataset - We should make sure to continue this conversation there to avoid fragmentation.

  • IA – ….

  • JW – Maybe forking liberally is the best way to keep everyone able to prioritize their own interestrs

  • IA – Would like to avoid duplication of effort and fragmentation of data

    • DD – Agree

  • DD – Do we have candidates for people who could lead a working group on this?

  • MH – That’s a tricky question. We’ll want to punt this to a meeting where we have all approvers present.

  • DD – Sounds good, we’ll bring this up again next week and get JC’s feedback - He may know folks interested in spearheading this. Questions to answer could be: Would there be a single leader or a committee, etc. Should think about this from a user’s perspective.

DD : current sprint - ends 4/3

• architecture overview : https://drive.google.com/file/d/1ZA-zuqrhKSlYBEiAIqxwNaHXvgJdlOkT/view?usp=share_link

• alchemiscale 0.1.0 milestone

• coordination board : alchemiscale : Phase 1 - MVP

• updates on In Review, In Progress, and Available cards

• DD – Alchemiscale 34 – Being closed by Alchemiscale 98 – This is close to merge and I’ll cut a release as soon as it’s in.

• IP – Perses 1066 – No progress. We wanted to meet with John to do the review and make sure that we’re testing what we want to test. I’ll schedule that meeting for next Monday.

  • DD – Perfect, this will be a really useful meeting for me.

  • IP – Yeah, and we can continue discussion of the results of that meeting here.

  • IP – IA, have you run a repex protocol for our test systems? I’d like to compare results.

    • IA – We have, but not yet with element changes. I’ll send you what we have

  • DD – Question for OpenFE folks – I want to try using the OpenFE BaseProtocol - Is it ready to use? I’d be interested to throw it into my testing.

    • RG – I’m working on it. Right now it just uses a single default setting. It should serialize OK. I’ve got this in a PR, will let you know once I’ve got this ready.

• IA – PLB 83 – Still waiting on JC.

• IP – Perses 1128 – We just reviewed the changes that JC made, and are trying to take work off his plate.

  • IA – Is this blocking testing of the Alchemiscale protocol? If so, we can pass the mapping directly by file.

  • DD – I’ve been testing using the OpenFE-benchmarks repo, which has been working for me. Also RG showed me how to use the OpenFE interface to LoMap.

  • IA – Those should already by generated by the objects?

  • DD – I don’t use the objects from OpenFE-benchmark, they’re broken. I opened a PR for this.

  • IA – Ah right, it got hung up with gufe was being changed. I’ll check out that PR.

  • IP – Are you running alchemiscale noneq protocol?

    • DD – Yes

    • IP – That may be a blocker - That’s OE dependent

    • DD – We can deploy with an OE license for now.

RG – OFEExamples 36 – It’s a little unstable. Needs other things resolved.

• DD – Anything I can help with? I’d like to share this with MOsato.

• RG – I could use your eyes on gufe PR 152. Will tag you.

• DD – Ok, I’ll check those out after 148 and 150.

JW – Any big possible risks we should beging preparing for? Do you hjve everything you need?

• DD – I expect early users to unearth a lot of bugs, and I’ve started preparing a series of milestones for subsequent releases with new features and bugfixes. 0.2 will be largely completing implementation of some minimally implemented features and bugfixes. 0.3 will be some small new features and optimizations.

• JW – This sounds great. It’s fine if we deviate from this versioning - There will be enough real work to do in all areas and so I like the grouping of tasks here but let’s not get hung up on versioning.

• DD – Also, thinking about branches and deployments - Could have different hosts for running prod, dev, and QA.

  • MH – Do we want all 3 to run on the same host?

  • DD – No, that’d be unnecessary work.

  • MH – Could mix hosts and have them swap jobs.

  • … (see recording around 40 minutes)

  • MH – I’ll set up the dev+QA box.

  • DD – It’s fine if we separate them into 3 hosts to keep things simple.

IA (chat) – I do have one extra thing - alchemiscale @ omsf workshop, are we interested in making it a day 1 talk? I think David Mobley was positive towards the idea, but it would be good to get it locked in (just so we can finalise (sic) day 2 things).

• JW – I’m in favor of this. It’s ultimately KCJ’s call but I’ll join up in recommending it. DD, are you up to give a 15-30 minute talk on alchemiscale on MAy 13?

• DD – Yes, this sounds great.

• .