/
2024-02-14 Meeting notes - OpenFE collab

2024-02-14 Meeting notes - OpenFE collab

 Date

Feb 14, 2024

 Participants

  • @Richard Gowers

  • @Matt Thompson

 Goals

  •  

 Discussion topics

Item

Notes

Item

Notes

 

  • MT: When is the grand refactor done?

    • RG: Probably tomorrow

    • RG: Will get CI stable on main repo, then migrate AFE Protocol into skunkworks

    • RG: will keep same directory structure so existing PR can migrate easily to target skunkworks repo

  • RG: This is how SolvationSettings pattern works:

    • RG: can define PackmolSolvation (or similar) as way to place packmol-specific parameters

    • RG: the code branch here can then be toggled based upon which XSettings object is fed into the AbsoluteSolvationSettings Use Interchange for some OpenMM object creation by mattwthompson · Pull Request #703 · OpenFreeEnergy/openfe

    • RG: by having this pattern we can test water solvation FEs back to back when doing end to end testing - these should be within statistical noise of each other

  • MT: unclear how some chemistry things are defined

    • RG: behold

  • MT: what about things like box sizes etc?

    • RG: This is how we are handling solvation in the future, this won’t be involved in this specific work project

  • MT: longer term beyond this project - how should combining OpenMMFFs products and Interchange products work?

    • both: this is necessary for use cases that don’t use rosemary, for example lipids/RNA/benchmarking of rosemary etc

    • RG: it’s not totally clear if add operations should work in openmm or interchange currency, both are fraught with peril

    • MT: correct

    • MT: ingestion of openmm to interchange is something that might become considered stable soon

    • MT: probably best way forward with this is openmm objects into interchange, combine interchanges, then re-export into openmm

    • RG: instead of dunder add, maybe interchange.add() function?

    • MT: yep

    • MT: Ok this seems like we’ll be able to handle the rosemary project (#3) later on with these assumptions

  • RG: let’s check how we’re progressing relative to the work plan

    • MT: currently vacuum usage in place

    • MT: will proceed with putting packmol usage into current PR

    • RG: cool, this can start getting merged into the skunkworks repo soon then

    • RG: let’s schedule to meet again in 2 weeks

    • RG: for future sprints, will have to write some scripts that generate AFE calculations so we can start trying out this code and shaking loose any issues





 Action items

meet again in two weeks
RG: to finish skunkworks migration
MT: to work on packmol usage & settings required

 Decisions