2021_08_13 Thompson/Wagner/Eastman OpenFF/OpenMM check in

ParmEd 1-4 scaling trickiness

• JW – When using ParmEd to combine components from AMBER prmtop (with 1-4 scalings attached to torsions) with components from OpenMM (with 1-4 scalings defined in openMM Exceptions / ParmEd adjusts), the 1-4s from the prmtop get ignored.

  • When the numbers started getting toward the correct range, I decomposed OpenMM nonbonded components such that I could match them with AMBER energy components (ES/vdW pairwise/1-4)

Making amino acid substructure dictionaries

• JW – OpenFF is looking to read canonical polymers from:

  • PDB-with-CONECT (elements and bonds), where we recognize substructures and assign bond orders and formal charges.

  • SDF, where we have elements+bond orders+formal charges, but need to assign canonical residue names

• We’re currently doing this by parsing aa_variants.cif from the PDB

  • Program to parse cif: https://github.com/openforcefield/openff-toolkit/pull/982/files

  • Example of “processed database” file https://github.com/openforcefield/openff-toolkit/pull/973/files#diff-cea06081529235c07b6eef4ba96e2ac35dd19809d3adc89e7fc70d7c8c035011

OpenMM release schedule

• MT – Plan for 7.6? Date/content?

• PE – Good to follow the forum. 7.6 beta is out, will build release candidate next week, full release likely a week after.

NIH interactions

• JW – Feedback if I hear back from the NIH? Gave this talk to them a week or two ago: https://docs.google.com/presentation/d/1yuZ7gSaP-fP101fZLfBoRk7GwtIvVYoNBOIzj5gpYs4/edit#slide=id.ge552d9a537_0_111

  • PE – Emphasize that they need to make research engineering a career track.

  • PE – Upstream OSS funding wouldn’t be super useful if it was re-decided every year.