Keiser Lab

Systems pharmacology and deep learning laboratory @UCSF

[Will Connell @wilstc]

🧬🔮 Single cell foundation models have been a recent hot topic in bio-ML! A few of the recent methods and some thoughts 🧬🔮

1) Geneformer
2) scGPT
3) scFoundation
4) Exceiver

[Keiser Lab @keiser_lab]

Congrats to the @vevo_ai team on joining the #AWSGenAIAccelerator cohort! Excited to support their groundbreaking work in #GenerativeAI, single-cell foundation models, & million-cell screens opening frontiers in drug design. Accelerator info: https://t.co/2uT0GEgZbD

[Nima Alidoust @nalidoust]

I'm pleased to share that @vevo_ai has been selected to join the @AWSstartups Generative AI Accelerator. Only 21 leading startups in the Generative AI space (out of 1200+ companies) were selected to be in the inaugural #AWSGenAIAccelerator cohort. (1/5)🧵
https://t.co/dVK5xdLKlk

[Keiser Lab @keiser_lab]

@fraser_lab @CyrusMaher @ORonneberger @danrsm @GarneloMarta @MichaelZielins @JonasAAdler @arkitus @CarlDoersch @pushmeet @yifan_ucsf let's do it

[Keiser Lab @keiser_lab]

Accuracy is not sufficient for using image analysis AI support tools in clinic. Our latest work by @AlbertY60806367 in @Nature_NPJ with the Wei Lab (https://t.co/NIdWOpnf7E) describes how computational "stress tests" help ensure AI models are clinic ready: https://t.co/pR4BkAtTVA

[Keiser Lab @keiser_lab]

@fraser_lab @lisa_gunaydin +1 (albeit untenured) https://t.co/hnthQ3tcMj

[Martin Kampmann @MartinUCSF]

@ianholmes @cziscience @czbiohub @Facebook Please submit your signature here: https://t.co/GwlvdxRNy9
we will update the letter on a rolling basis as more CZI-funded scientists sign

[Keiser Lab @keiser_lab]

As a lab funded by the awesome @cziscience, we stand with our colleagues in an open letter to Mark Zuckerberg asking @Facebook to enforce its own policies on misinformation and incendiary statements in this critical time

[Martin Kampmann @MartinUCSF]

My lab is funded by @cziscience & @czbiohub, awesome organizations whose values I share. Many CZI-funded scientists are concerned about the spread of misinformation & incendiary statements on @Facebook. We wrote an open letter to Mark Zuckerberg:
https://t.co/XzDoac9YRp

[Keiser Lab @keiser_lab]

Congrats to @kangway, who joins the 2019 class of Arnold O. Beckman Postdoctoral Fellows in Chemical Sciences! Thank you @BeckmanFnd for the game-changing support & recognition of an exceptional scientist.

[kangway chuang @kangway]

Thank you @BeckmanFnd for the opportunity to go after exciting new research directions! Feeling unbelievably excited and grateful; all my thanks to @keiser_lab and @IanSeiple for their mentorship @UCSF , and again to @sarah_reisman for making me the chemist I am today!

[kangway chuang @kangway]

Wow, I'm so impressed by @UMich undergrad @YannahMelle, who just gave an awesome talk on her summer research in the @keiser_lab as part of the UCSF Summer Research Training Program! Great job and thanks so much for making the lab awesome, can't wait to see what you do next!

[Keiser Lab @keiser_lab]

A #redteam finds fragility in our security models. A scientific #AIredteam would find where our #machinelearning models, #databias, and assumptions are just as fragile. #reproducibility

[UCSF Bakar Computational Health Sciences Institute @UCSF_BCHSI]

Scientific machine learning - @Keiser_lab proposes scientific AI red teaming to find ways if our assumptions are wrong. Does your model make scientific sense? #UCaiBio

[Keiser Lab @keiser_lab]

If a #datascientist is a scientist, shouldn't they also be a #databreaker?

[India Hook-Barnard, PhD @SelectAgent]

“Let’s move fast and break our own things.” @keiser_lab calls for scientific method, responsibility, red teaming for AI. #UCaiBio

[Keiser Lab @keiser_lab]

We actually hit 2400/hr -- about two annotations per standard human breath!

[Atul Butte @atulbutte]

Brittany Dugger: With gamification, easier interfaces, and rewards, a pathologist can annotate 2000 images an hour! #UCaiBio paper here: https://t.co/tUHYm01DHr

[Atul Butte @atulbutte]

LOVE the systemwide #AI collaborations already happening within the @UofCalifornia! https://t.co/tUHYm01DHr #UCaiBio

[Dr. Camille Nebeker (@camillenebeker.bsky.social) @cnebeker]

Looking forward to moderating our #UCaiBIO panel on #Responsible Use of AI in Biomedicine” 6/18 with @keiser_lab @SciTechProf @LaurelRiek !

[Keiser Lab @keiser_lab]

Want to try your own #DeepLearning models for #AlzheimersDisease pathologies? Ziqi Tang's paper, code, AND data are open to everyone - let us know what you discover! https://t.co/RELEPBL2JF - This has been a fantastic collab w Prof Dugger @ucdavis.

[UCSF School of Medicine @UCSFMedicine]

“It’s a co-pilot, a force multiplier that extends the scope of what we can accomplish and lets us ask questions we never would have attempted manually," explained Professor Michael J. Keiser, PhD, of @INDUCSF & @ucsfpharmacy https://t.co/mPnMjfvnT9

[Keiser Lab @keiser_lab]

@jeremyphoward @danaludwig Good eye @jeremyphoward! This wasn't scoped as an online course, so some notebooks contain unpublished/nonpublic examples/data. But I do have a couple ideas on what we might be able to share - DM me

[Keiser Lab @keiser_lab]

This year's "Advanced Topics in #DeepLearning for Biological Sciences (BMI 219)" course starts Monday morning! Looking forward to Dr. Nobuyuki Ota's third year of cutting-edge #ML lectures - #UCSF https://t.co/r8sfPU0BGW

[Preferred Networks @PreferredNet]

Nobuyuki Ota, COO of Preferred Networks America, lectured on "Deep Learning for Biological Research" at the University of California, San Francisco from April 23 to May 11.
This course introduced students to current applications of deep learning in biological research.
#UCSF

[Keiser Lab @keiser_lab]

@iRonB25 @atulbutte These aren't recorded, but if we could show there's enough interest we may be able to get the logistics/support for it next time around

[Keiser Lab @keiser_lab]

@nxpatel @nxpatel DM me & we'll see what we can arrange!