2022-10-20 Protein FF meeting note

ForceBalance fits for proteins

@Chapin Cavender

• (Slides to be uploaded)

• MS - The weighting scheme doesn’t seem right. I might be losing the logic.

  • CC to clarify later

Check in on status of protein validation livecoms paper -

@Michael Gilson @Chapin Cavender

• would be good if published by grant renewal time.

• @Michael Shirts completing it will be a priority for me starting November (following Nov 5 R01 due date!)

• CC: status

  • have a complete draft from authors, currently going through it

Review of blockers for having a test suite prepared and running it so the team can help Chapin get rid of them

@Chapin Cavender

• Also, what would need to be done to run espaloma proteins through at the same time?

  • CC – no scientific or conceptual blockers there, just time; hopefully will have more after forcebalance fits are done. Hopefully fits will go smoothly now

  • CC – as long as we can generate OpenMM system from espaloma protein, we should be able to run benchmarks the same way

  • CC – maybe can run espaloma benchmarks on lilac or similar. Will reach out to Yuanqing at suitable time

  • MS, MG, CC – can reach out to YQ while protein benchmarks are running, give some advance notice

• Time estimates

  • CC – run guesstimated at about a week

  • Fitting ff14sb and null model – results possibly out in a week or two

  • PB – what’s the longest torsion scan in the fits by time?

    • CC – about an hour