2021_10_06 Pulido/Wagner check in

Perses/OpenFF/OpenFE coordination

• JW – We’re trying to figure out how OpenFF, OpenFE, and perses devs can meaningfully work together. Still a lot of unknowns. We had a meeting this morning that I thought about inviting you to, but it would have been really last-minute. But I can loop you in for future meetings if you want

• IP – Please loop me in.

TopologyMolecule deprecation

• • JW – Diff is too large to meaningfully compare. But I’ve noted a few places in particular that exhibit the major changes in the PR.

  • IP – I’ll skim over this and give feedback.

Residue perception/performance improvements from splitting at peptide bonds.

• IP – Had limited time last week since I had problems with OpenMM. Some low-level CUDA things with how openmmtools context cache was possibly leaking memory.

• IP – Tried to speed up reading-from-PDB-and-assigning-bond-orders by splitting at peptide bonds. First attempt was to split the molecule at peptide bonds by perceiving the simple subgraph.

  • (working session, found problem with histidine SMARTS and patched, will continue working tomorrow)