what can a deeply phenotyped human cohort tell us, and what can today’s AI models do with that information?
we built PhenoBench around the Human Phenotype Project: 90 tasks across 15 clinical domains, using clinical, imaging, molecular, and wearable data.
having those tasks let us ask a few things: which measurements actually help? and do more capable models make better use of them?
on these tasks, tabular foundation models performed better overall, but the gains over ridge regression were small. the LLM results felt like a useful example of the “jagged frontier”: they did relatively well on some tasks and poorly on others. models trained on the same input fields generally did better.
btw, most of the work was figuring out what to ask and how to tell whether a model’s answer was any good. that took much longer than running the models. having done that, trying another model or measurement is a lot easier.
Last week we've gathered 80 colleagues from the biotech industry to discuss how they're integrating AI agents into data teams.
We wrote seven quick takeaways on how agents are already transforming biotech >>
Sharing Gluformer, our latest paper @Nature: The first generative foundation model for blood glucose data, trained on data from 10,812 adults of the Human Phenotype Project
Gluformer predicts risk of diabetes and cardiovascular outcomes better than the standard of care
Full paper: https://t.co/SrdxDzftYU
Collaboration of @WeizmannScience@mbzuai@nvidia@Pheno_AI
Work by @GLutsker Gal Sapir, @smadarshilo, Jordi Merino, @nastya_godneva, Jerry R. Greenfield, Dorit Samocha-Bonet, Raja Dhir, Francisco Gude, Shie Mannor, Eli Meirom, @ericxing, Gal Chechik, and @H_Rossman
An impressive sleep study goes deep on links to lifestyle and diseases
—impact of gut microbiome and diet re: sleep apnea
—big contribution of lifestyle factors
—ability to predict across 16 body systems from sleep data
@NatureMedicine@segal_eran@WeizmannScience
https://t.co/2ifQvV043Z
🎉 Thrilled to share that my latest paper (w/ @segal_eran) on the gut microbiome's role in host metabolic health has just been published in @NatureComms!
Dive into our insights📄✨
https://t.co/WhrNfuLN7o
1/5
Thrilled to share our work, GluFormer: A foundation model for continuous glucose monitoring (CGM) data, trained on over 10 million measurements from over 10,000 individuals.
Read the full paper here: https://t.co/pWt8IXRH7y.
1/8
Hundreds of Jewish students at @Columbia just published one of the most incredible student letters I have ever read.
It's not only magnificently written, but it also clearly articulates their experiences on campus for the past six months.
Their letter tells the story of what's it like being a Jewish student right now better than any professor like myself could ever do.
Please take 4-5 minutes to read their letter.
Give Jewish students a voice.
https://t.co/nsWNScaoZS
🙌 I've had the privilege of leading this effort at @eleventx alongside @oweissb, Ohad Yogev, Adi Naamati, @CMordstein, Owen Cox, and many others! and I hope the 𝐱𝐑𝐍𝐀 𝐃𝐚𝐭𝐚𝐛𝐚𝐬𝐞 (built on @NotionHQ) will empower Eleven and its therapeutic molecules for years to come 🌠
Every biotech struggles with cataloging and tracking its molecules and their experimental results. We wrote a blog post about using @NotionHQ to seamlessly design, catalog, and track our programmable molecules (link in next tweet).
Great work by @AnatEtzionFuchs!
Excited to share our work on an ELN that serves us instead of us serving it, using @NotionHQ !
It started as a cost-saving move, but turned into a major restructuring of how we organize and align the company around shared work processes and mental models
https://t.co/XVXT9ARo27
Business as usual @BBC. As if there aren't any Israeli hostages held by Hamas, and as if Hamas isn't responsible for the life of Gazans. 8 minutes of blaming israel for everything wrong in Gaza, without mentioning Hamas. Not productive.
As a @Princeton graduate I'm deeply disappointed by not hearing a clear condemnation of the terror events committed by Hamas, from Ivy leave universities. If you are "with both sides" you clearly and openly support terrorists!
#HamasisISIS
Imagine a 9/11 with 10x times more murdered; with many more children and elderly massacred; with the terrorists conquering US towns and having the choice to spare each of the fatalities separately, but again and again choosing to butcher them all, sometimes also raping them; with the terrorists also desecrating the bodies of the murdered, and parading them, sometimes naked. Think about how you would feel. Think about how you would feel if on that day people asked you to consider both sides. Think about how you would feel if people said their thoughts are "with both sides on this complicated day." Think about how you would feel if people that day called AlQaeda "freedom fighters."
What you would feel in this hypothetical situation is what Israelis feel right now. What I described is only part of the atrocities that Hamas Palestinian terrorists committed in their attack on Israel yesterday. (Israel's population is about ~33 times smaller than the US, hence the approximate number 10x so far.) And I haven't even started to talk about the kidnapped men, women, elderly, and children, whose fate I cannot bear to imagine. (And, I'm sparing you from some things that you would never be able to unsee/unimagine.) As you read and hear various "thinkers" and journalists in the upcoming days and weeks, I urge you to remember this post. Sometimes in life the truth is not in the middle. Sometimes in life, there is nothing more immoral than claiming that the truth is somewhere in the middle.
This amazingly accomplished panel @WeizmannScience discussing challenges Israel faces in becoming an impactful #RNAtherapeutics hub: lack of chemists, not enough experience in scaling up. And the edge that we have: innovation, out-of-the-box thinking, and a great eco-system
Israeli RNA tweeps - our meeting on RNA therapeutics is at the @WeizmannScience campus on the coming Monday. Agenda here, exciting speakers and panel. Still possible to register (for free): https://t.co/tIeqVPzNWT