Page Comparison

...

Updates from MolSSI
- Ben’s introduction to new features in qcportal
New submissions
- CC’s `OpenFF Protein Capped 1-mer Sidechains v1.0` : 46 2D torsion scans
- Modified submissions: SPICE sets v1.2 (pubchem sets in queue)
Throughput status
- OpenFF ESP Industry Benchmark Set v1.0: 1327014145/56054 ~24% ~25% from 12% last week
- Openff dipeptides torsiondrives v2.0: 24/26 TD complete from 23/26 last week.
  - Stuck here, no change in # of optimizations for the last five days. Errors in scf convergence.
- SPICE DES370K Single Points Dataset v1.0: 99% Complete! 3637 MBIS failures, moved to end-of-life.
- SPICE Dipeptides Single Points Dataset v1.2: 2251124321/33K ~ 67% 71% from 37% done
User questions/issues
Science support needs
Infrastructure needs/advances

\uD83D\uDDE3 Discussion topics

...

...

Item	Presenter	Notes
	David Dotson	DD – There’s been an update to our production environment, please restart your managers. I’ve posted instructions in qcfractal-compute. CC – Will do
Upcoming improvements preview	Ben Pritchard	QCFractal/QCPortal new features Datasets are matrices - Rows are entries and columns are specifications. A particular intersection of a row and column is a “Record” (though because it’s a database, we have a linking layer, where the object is called a “RecordItem”) … BP – For dataset retrieval, would fetching a pandas dataframe be useful? CC – So, like, I could fetch all the final energies? BP (shows an example of how to do this around the 35 minute mark of the recording) PB – Could I modify a submission? BP – Not really, you’d have to talk to GeomeTRIC JW – Can we look up molecules by smiles/metadata? BP – Yes, the metadata can be fetched more easily (~38 minute mark in recording)