💊 Can AI combat the opioid epidemic? In a paper recently published in @INFORMS, @mit researchers introduce a ML model that detects and flags suspicious opioid buyers engaged in improper purchasing behavior, w/ an AUC of 0.87. https://t.co/36WTYFGI1U

#JameelClinic AI faculty lead @BarzilayRegina predicts that AI adoption will be significantly faster in countries w/ developing economies. "They have serious unmet needs and the clinician understands they need other help." @NPRGoatsandSoda @gabrieman https://t.co/njRenjdrF0

🫁If CT scan features can predict lung cancer, their Mutual Information (MI) is high. A #NeurIPS2025 paper by Reuben Dorent, @GollandPolina & William Wells bridges MI and JSD, helping AI models learn more meaningful patterns. https://t.co/eedSdtWvdY

📸 Memories from #MoML2025! We just sent out a short feedback survey to #MoML2025 attendees. As a thank you for completing the survey, you'll be able to access professional photos taken at the conference!

Molecular modeling has just leveled up! 🤖 BoltzGen, a new generative model created by Mark Foundation grantee @BarzilayRegina & her colleagues at @AIHealthMIT, can generate new protein binders against any biomolecular target.  https://t.co/5ZymOutADZ

standing room for a research seminar in mit's largest lecture halls. ive never seen anything like it. congratulations to the boltz and boltzgen teams for this true tour de force.

“We don’t just test them on computational metrics, we test them in the real world…and on hard problems.” @HannesStaerk on BoltzGen

BoltzGen seminar has begun! Incredible turnout, thanks to everyone who showed up despite the rain and wind! ☔️🌬️

If you haven't registered already, don't miss out on the chance to hear about BoltzGen straight from @HannesStaerk, @BarzilayRegina, and Tommi Jaakkola tomorrow @MIT from 5pm to 6pm at Stata Center, which is open to the public! 🗓️✅RSVP now: https://t.co/YL0cWTwoI2

BoltzGen Democratizes AI Therapeutic Design, Expands Druggable Universe The open-source model designs therapeutics across "any" modality with experimental validation for diverse real-world challenging targets @dreidco #AI #drugdiscovery #proteindesign https://t.co/jTWh9jQJRW

Congrats to #JameelClinic PI @dina_katabi and our new Advisory Board member Ravi Thadhani on being elected to the @theNAMedicine's Class of 2025! 🎉 Read more: https://t.co/ZvERFfUDTv

Join us in-person at MIT on Oct. 30 @ 5pm to hear from @HannesStaerk on BoltzGen and the state of biomolecular design! 🗓️RSVP now: https://t.co/RtPuWLCDby

Excited to release BoltzGen which brings SOTA folding performance to binder design! The best part of this project has been collaborating with many leading biologists who tested BoltzGen at an unprecedented scale, showing success on many novel targets and pushing its limits! 🧵..

Thank you MoML! I will spend the novelty cheque wisely

Congrats again to @NYU_Courant’s @AlanNawzadAmin for winning the Best Paper Prize for his work on a diffusion model called SCISOR, which is trained to make deletions that arrive at a natural protein sequence whilst maintaining protein function! #MoML2025

A big congrats to @NYU_Courant’s @AlanNawzadAmin on winning the prestigious Octavian-Eugen Ganea Prize for Best Paper for his work “Shrinking Proteins with Diffusion”! 👏🏆#MoML2025

Last but not least in the Top 5: Ruijiang Li of @SanofiUS presented his work on SNAC-DB, an ML-ready database and pipeline that can help accelerate gains in accuracy and generalization. #MoML2025

Next up in the Top 5: @joh_sweh presents mechanistic interpretability for pair representations in protein co-folding! #MoML2025

Part II of the Top 5: @KevinKaichuang from @MSFTResearch presents the Dayhoff Atlas, datasets and models for protein generation. #MoML2025

Up next: what IS inverse folding in protein design and what ISN’T inverse folding in protein design? @MIT professor @sokrypton breaks it down @ #MoML2025