SF AIxBio Reading Group #5
This week, one of Biopunk's very own Nina is gonna host journal club! Nina is CTO and co-founder of FlowState (@flowstatebci), building wearable neurotech that read your brain state and adjust audio in real time to help you enter deep work and the flow state. Sign up for our waitlist here: flowstatebci.com/#waitlist
Guiding question:
Can you skip the physics? Generative models that emulate protein motion instead of simulating it.
Resources:
Main paper (free preprint): Lewis et al., "Scalable emulation of protein equilibrium ensembles with generative deep learning." https://www.biorxiv.org/content/10.1101/2024.12.05.626885v1
Published version (Science 2025, paywalled): https://doi.org/10.1126/science.adv9817
Optional, 5 min: the code and a runnable chignolin example, https://github.com/microsoft/bioemu
Optional, for the kinetics thread: "Accelerated sampling of protein dynamics using BioEmu augmented molecular simulation" (bioRxiv, Jan 2026), https://www.biorxiv.org/content/10.64898/2026.01.07.698041v1.full
No background needed. If you've never touched molecular dynamics, read the abstract, Figure 1, and the discussion, and come anyway. The interesting question here is epistemological, not technical.
How to prepare:
Read / skim the paper.
Bring an answer to the guiding question, or a question you'd like to ask the group.
Please bring your own copy of the material (physical or digital).
Arrival instructions:
I'll be at the front door to let you in. Please have your QR code from Luma so I can scan you in!
By RSVPing to this event, you agree to:
Help me foster a community where we put our biases and egos aside and learn from each other.
Respect the no tolerance policy for misbehavior.
Believe that no question is too "small" or "stupid."
Contribute to the discussion. I'm keeping this group small so we can all have a chance to share, and I want to hear from you.