Pontibus handover

Participants

Goals / Agenda

• Summary of context, purpose and implementation of pontibus

• Usage

  • Who will be the end user?

    • LW: Interested to hear IA’s thoughts as maybe I misunderstood, but I assume we are the end-users of this protocol specifically, and then affiliated scientists would be end-users of any product we create from it (e.g. tying into Evaluator)

  • How much work is it to get them running?

    • LW: OpenFF anticipates running benchmarks to compare results to previous workflows (which will differ slightly as they, e.g., used the yank trailblazing algorithm). Is there anything we can build off easily to do this, e.g. any previous tests or benchmarks OpenFE has run, possibly of Sage 2.0’s FreeSolv and MNSol data?

    • LW: a good target for OpenFF would be to try to get things running in the week following this meeting, and come back with issues/questions in a second handover meeting.

  • How much / what compute will they use

• Maintenance

  • Will this become an OpenFF product?

  • Will it move to an OpenFF repo?

    • LW: in general who should take ownership of this – if a PR is opened, who is expected to review?

  • How will we manage maintenance?

    • LW: how much does OpenFE want to be involved with pontibus going forward?

    • LW: This meeting is for the v0.0.1 handover but there is a v0.1.0 milestone, does OpenFE/IA anticipate working through these issues here or should OpenFF?

    • LW: pontibus is built on openfe/gufe; of course this might be difficult to forecast, but is there an estimate of the expected rate of change upstream that would necessitate updates to pontibus?

  • Does pontibus depend on openmmforcefields?

•  

Action items

Decisions