Timothée Keyes, PhD

In the spirit of being an MD/PhD student for the rest of my life, this week I started a new job as a data scientist on Stanford Health Care’s AI team! I’ll go back to clinic eventually, but in the meantime here’s a cartoon that DALL-E made to commemorate the occasion.

@elliotoxxo I was shocked at how true this was when I moved here. Welcome to the bay!

Just landed in San Francisco and it is so cold don’t yall know it’s July ???

🧹🤩 #tidyomics made it to @naturemethods! "The tidyomics ecosystem: Enhancing omic data analyses", 👏🏽, @hutchisonwj0, @timothykeyes, and Team! 32 researchers | 26 institutes | 10 countries | 4 continents A #crowd/#community-research success nature.com/articles/s4159… @cziscience

🚀 Checkout our ICLR 2024 Spotlight Poster #165 today (Session 3) 🩺 "MOTOR: A Time-to-Event Foundation Model For Structured Medical Records" ✨ Highlights: - The first TTE foundation model for structured EHRs - Improves SOTA TTE by 4.6% & boosts label efficiency up to 95% - TTE pretraining scales to 16k tasks & reduces GPU memory usage by ~35% - Model weights available for research use! 🔍 Tutorial: github.com/som-shahlab/mo… 🤖 Model: huggingface.co/StanfordShahLa… #ICLR2024 #AI #Healthcare

Really interesting 🧵 about evaluating large language models’ performance in a health care context!

💊Clinical trials cost $100M’s, yet 1 in 3 fail due to insufficient patient enrollment.🚨Despite costing $15B/yr, patient recruitment is a key unsolved bottleneck in drug discovery. Our latest work shows how LLMs can help solve this! 🌐Site: clinicaltrialmatch.stanford.edu (1/7)

Yesterday, @JCI_insight pubbed our study examining MD-PhD program acceptance & how it varies by demographics & research experience. 1st-gen, Black or reapplying applicants were less likely to be accepted, regardless of research experience & publications. insight.jci.org/articles/view/…

shoutout to everyone else who had a really weird phd experience. as long as you tried your best, that’s what really matters. so much of it is out of your control, be kind to yourself

defending my phd was wonderful and exciting and terrible and gut-wrenching all at the same time and even though i can’t decide if i feel happy or absolutely dismal right now, i know i did something really hard and that’s an accomplishment to be proud of

The lack of a unified electronic health record system in the US is a threat to public health and absolutely affects patient outcomes

@MikeGisondi @StanfordMed Thank you!!! I have been asleep for the past 16 hours

As soon as I got my hands on GPT4, I challenged it with multiple fraught clinical questions and scenarios to see how I could break it. I asked it to role-play counseling me in the way I had to counsel a real patient's family member through a difficult situation.

@RoxanaDaneshjou i haven’t tweeted in months so I think it’s safe to say that the algorithm has abandoned me lmao

@timothykeyes This tweet should be doing numbers.

scientists are really out here writing “data availability” “code availability” and “statistical analyses” sections like they’re playing Two Truths And A Lie and i am losing my mind

🎉🧹 The #tidyomics ecosystem is official! Into #omic data analysis? Spanning #Seurat @Bioconductor #SCE, #SE, #GRanges, #Citometry? Now, just use #tidyverse ! Co-led with @mikelove and @TonyPapenfuss @WEHI_research #singlecell The preprint biorxiv.org/cgi/content/sh…

@cycl0pentadiene yo do you really

today i wrote a jupyter notebook that called an R package that called a python package

I am so proud of the rich history and the future direction of our #CancerBiology Program. Happy 45th, crabs! bit.ly/47Jy0b0 @stanfordcbio @AttardiLaura @SageJulien

(i just made a LinkedIn)

thoughts from the job market: i don’t want to “sell myself.” i want to talk honestly about my skills, interests, needs, and goals - and for that to be enough!!!