2026-01-28 All-Hands Meetings

Project Updates

• Alchemistry and OMSF workshops will be in Barcelona in May

• Alchemistry May 4-6 @ UPF https://omsf.io/alchemistry/  

• OMSF Symposium (incl OpenFF annual workshop) May 7-8 https://luma.com/of677kui  

  • BM – hotel recommendations?

  • DM – Hotel Barcelona Colonial

  • DC – will OpenFF get any breakout time to discuss our project? That would be valuable to me.

  • JW – We can’t really take an extra day since the Alchemistry workshop is before and I believe there’s a staff event immediately afterward. Maybe we can request a breakout room during the OMSF symposium.

  • MS – the sooner we can decide on this, the better I can make travel arrangements.

  • JE – I will talk to Zach from OMSF to see if a breakout session is possible.

• Chris Iacovella is joining us from the Chodera lab on a half-time basis

Science Updates

Recording: https://youtu.be/3q7_0l8pPoQ

Slides: https://docs.google.com/presentation/d/1HLtDh96WKgNNTklN_p-jqBFKKmqRRq1ZsbjUBADPHK8/edit?usp=sharing

• DC – any further insight into the protein force field work that’s going to involve espilon refitting of LJ parameters?

• 
  
  

• JC – Not positive, but I think it was that fitting to J couplings NOEs was mangled by longer-distance couplings also being in the fit. Implied that there was a medium range parameter that needed to be fit in addition to the torsions themselves.

• CC – Fitting protein-specific vdW parameters is out of scope for Rosemary-3.0.0. The goal is to do a first-pass fit of epsilon to see how much changing epsilon impacts NOEs to understand whether it makes sense to continue fitting torsions to NOEs.

•  

• .

Infrastructure Updates

Recording: https://youtu.be/huA4m_c_tEw

Slides: https://docs.google.com/presentation/d/1uc9W7_WHGUtRoa8aMLzgD1HovoKiU4bMERR6CsDYf2A/edit?usp=sharing

Pablo documentation: https://openff-pablo.readthedocs.io/en/stable/

• CC – is anonymous matching in pablo on by default, or does it need a flag?

• JM – by default, the list is empty so there’s nothing to match. You have to give it the molecules you want to be able to identify.

• CC – A few months ago I had looked into using pablo but at that point it wasn’t mature enough for production. What do you think now?

• JM – It will probably work. If y our existing workflow is painless enough, you don’t have to switch over. But I expect pablo would meet your needs.

• JW – Pablo is tested pretty thoroughly so I believe it will not give you incorrect structure loading.

Shouts-outs

• JE – Shout out to staff for staying in focus mode and getting things out the door.

• JW – Matt has basically been a fire-jumper this past quarter. We have a lot of complex, inter-depndent collaborations with other organizations. Each meeting has very high cognitivie overhead. And Matt has been jumping into these meetings and handling them admirably.

• MS – Shout-out to Lily for pushing the AshGC paper through!

• JW – Shout out Chris for getting into the standups with us and “believing in Santa Claus” with us.

• CI – It’s been pretty easy to integrate into the “weird workflows” and everyone has been really helpful in onboarding, setting expectations, and answering questions about how to work in this system.

• CC – Shout-out to Lily for helping me run the Tier 3 protein benchmarks. She tapped into a lot more compute power than I had access to at USD.

• LW – Shout-out to the MSKCC IT team for making all of this possible.

Q&A

• JW – FYI, changes upcoming to target density behavior in pontibus.

• JM & LW – (thumbs-up)

• CC & CI – not using pontibus at the moment, so if we need it later we are aware that there’s something to check on.