2022-02-22 QCA Submission meeting note

Participants

• Jeffrey Wagner

• Pavan Behara

• Ben Pritchard

• David Dotson

• Willa Wang

Goals

• Updates from MolSSI

• New submissions

  • Pubchem set1 local submission in process (Thanks to David!)

  • Modified submissions: SPICE sets v1.2 (other pubchem sets in queue)

  • SPICE DES370K Single Points Dataset supplement v1.0

    • PB – Should dataset specification be renamed from spice-default to spice-no-mbiis? This is just a subset of the original set, containing only the failed calcs.

      • DD – I think it’s a good idea to give it a different name from the default.

• Throughput status

  • OpenFF ESP Industry Benchmark Set v1.0: 16170/56054 ~28% from 25% last week

    • pretty slow, need to prioritize over spice sets on Trevor’s managers

    • PB – In the first week, this went from 1% to 25%, then this week it only gained a few more percent

    • DD – I don’t have any workers currently running with this tag - Currently I have workers with different tags submitted but the openff tag workers are lowest priority.

    • PB – We should prioritize compute managers in the order of openff, then openff-tscc, and then openff-spice at the lowest.

    • DD – Ok, I’ll put openff as top priority on PRP

  • Openff dipeptides torsiondrives v2.1: 24/26 TD complete

    • TDs with geometry issues resubmitted with corrected initial conformers

    • There is an uptick in # of optimizations, so it seems to move along the right direction

  • OpenFF Protein Capped 1-mer Sidechains v1.0: 0/46 TD, 60K optimizations from last week

  • SPICE Dipeptides Single Points Dataset v1.2: 33648/33K ~ 99% from 71% done

• User questions/issues

• Science support needs

• Infrastructure needs/advances

Discussion topics

Action items

• David Dotson will set openff as top tag on all PRP and Lilac deployments
• @

Decisions