2021-07-09 QC meeting notes

Hanging INCOMPLETEs on OpenFF Sandbox CHO PhAlkEthOH v1.0 OptimizationDataset

Ben

• BP: remedied issue, cleared and COMPLETE on all records

  • found race condition that likely produced this:

    • manager started up, and in the process of receiving tasks, immediately shut down; put the tasks into a broken state

MP2 specs

Pavan

• Added new compute spec to Openff Theory Benchmarking Set v1.0

New psi4 version for production

Pavan

• qca-ds-sub#214

  • want to up deployed psi4 version for iodine support

  • will need latest QCEngine 0.19.0

    • DD: last I tried this with openff-benchmark, hit issues I didn’t understand at the time; will check that this works at least for our deployed managers

• PB: tangential to this, worked on the questions Thomas Fox raised of certain chemistries being modeled poorly by the FF. Got good structural reproduction, so not clear this is the problem.

  • DD: feel free to engage directly with Thomas on your results; would definitely help to try to identify what else could be causing the geometry issues