DC has 80000 sugars optimized at B3LYP / def2-SVP, would like to use those structures as starting points for single point calculations at B3LYP / def2-TZVPP
DM (not on call) suggests considering keeping the level of theory consistent with other OpenFF, but running both to see which is better long-term (“asking, not telling”). Largely questions for open QM role
DD: Does QCA support both basis sets?
BP: Shouldn’t be a problem, as long as the basis set is available in Psi4 (which is should be)
DC: TZVPP basis set has roughly twice as many basis functions as DZBP (roughly 1000 for one disaccharide compared to 500)
DC Has existing data in orca format, DD directing toward qca-dataset-submission repo and helping submit PR
BP: Similarly-named basis sets (3 that sound like DZVP) may actually be in different families
BSE to c-f
MT: Basis Set Exchange migration to conda-forge started, need a PyPI release to package license file. BP working on some other issues in the package, new release soon
0 Comments