Meeting structure and agenda

Owned by Karmen Condic-Jurkic
Last updated: Feb 05, 2022 by Karmen Condic-Jurkic
2 min read

Proposal: https://openforcefield.atlassian.net/wiki/spaces/OFFO/pages/198672620

Big picture – “ideal solutions” – creative brainstorming for science + infrastructure

Meeting notes: https://openforcefield.atlassian.net/wiki/spaces/MEET/pages/201850984

Concrete plans

Force fields [45 minutes or less]

  • Feature specification for the next two small molecule force field releases

    • Essential features

    • Optional features

  • Feature specification for the first protein force field/release of first protein FF

    • (Simmerling plan – refit protein-specific torsions using data used in recent AMBER protein FFs, but with our bonds/angles/valence).

  • FF generation vs release – naming scheme clarification

Key infrastructure components for next 1 yr [2 hr or less]

  • Infrastructure overview [20 min or less]

    • Data management

    • Automated molecule submission

    • WBO support

    • Benchmarking dashboard/automated benchmarking

    • Python system object

    • …?

  • Prioritization of infrastructure [ 40 min or less ]

  • Details of infrastructure – define inputs and outputs, features, restrictions, timelines [1 hr or less]

Supporting science [1 hr or less?]

  • NIH planned roadmap:

  • Science topics:

    • FF optimization = chemical perception + properties + ForceBalance

    • ForceBalance vs Bayesian framework

    • Charges / Electrostatics

    • WBO parameter interpolation:

      • Basic science is in place

      • Someone needs to begin getting into FFs

      • Someone needs to push science forward

  • Benchmarking | Diagnostics – what do we have and what do we need?

    • QM data → conformers (geometries, energies, torsional fingerprint deviation (TFD), ligand scoring, etc)

    • Properties → feasibility study, data selection and curation

    • Protein-ligand --- Host-guest (proposal – use smaller HG systems for development of benchmarking framework, including analysis)

    • Biomolecule structural/energetic data (NIH grant planning)

      • Protein

      • DNA/RNA

      • Lipid

      • Other (sugars)

      • Convene our SAB to collect data. Maybe write a LiveCoMS article?

    • Dashboard – what data will be displayed?

  • QM level of theory – is it a settled thing or is another benchmarking study needed?

Hiring needs in next year

  • We have about 1 person/year off of NIH funds (maybe 1.5), after hires of Thompson/Dotson/Cerutti.

  • What missing expertise do we need?

Creation of OpenFF affiliates

  • Reward people who aren’t PIs who contribute:

    • e.g. How can Dennis Della Corte contribute (We discussed parametrization of radicals, or the development of additional tutorials).

      • He is also writing a grant for force field development for molten salts, which will fund a graduate student who has professional programming experience and can partly support the Dev-Ops aspects of OpenFF.