Posts on X

Daniel Cremers @TU_Muenchen came over from his sabbatical @UniofOxford to present not only his research but also our Munich Center for ML, big thanks. #baiosphere

Bastian @Pseudomanifold moderating our morning ai.bay 2023 session on ChatGPT & friends: great discussion by Verena Rieser @EdinburghUni, Hinrich Schütze and Barbara Planck @LMU_Muenchen #baiosphere

Daniela Rus @MIT_CSAIL beautifully summarizing technical challenges of AI at ai.bay 2023. #baiosphere

Alex @smolix talking at ai.bay 2023 on ML teaching, AutoML, and foundation models. Exciting! Plus, he just started boson.ai- congrats! #baiosphere

Anna Bauer-Mehren @Roche_de talking about AI in healthcare at ai.bay 2023. #baiosphere

Bavarian PM @Markus_Soeder talking at our ai.bay 2023 conference about AI as the ‚warp engine’ for new developments

Our Bavarian AI conference being kicked off by Gitta Kutyniok, @ja_schnabel & Angela Schoellong. Julia: AI in health is key not only for diagnosis but also prevention. #baiosphere

Congrats to @morris_lab for CellOracle out @Nature today! Joining recent approaches (Pando, Scenic+, chemCPA) to aim at OOD prediction of perturbation effects, CellOracle builds a putative gene regulation model, and has an impressive set of validations. nature.com/articles/s4158…

"Focusing is about saying no" - Steve Jobs (WWDC'97). You might have seen this, but still gets me wondering if in academia our one-student-one-paper culture is making us do too many things? youtube.com/watch?v=H8eP99…

Excited to finally see ExpiMap out @NatureCellBio - led by @Mohlotf & Sergei Rybakov, we inform single-cell embeddings by pathway priors (+ newly-learnt ones). This allows for biologically understandable components in the latent space and program queries. nature.com/articles/s4155…

Lovely @xkcdComic on K-means clustering.

Finding a key zoom feature after 3 years: you can quickly copy invite link via shift-command-i or little up arrow next to participants after starting new zoom session. Nice.

Physics-informed deep learning has been getting quite some attention in physics (besides ML), see eg. nature.com/articles/s4225… - what are (or could be) examples of such constraints and models in cell biology? Thinking population models, RNA velocity, signaling pathways etc.

Interest @naturemethods paper by Qing Nie lab @UCI_CMCF that allows to learn cell–cell communication from spatial transcriptomics data based on optimal transport. nature.com/articles/s4159…

Opening event of our new @TU_Muenchen School of Computation, Information and Technology.

Lovely single cell intro by @Dominik1Klein at lab meeting today

Big New Year’s Reception with our Bavarian prime minister @Markus_Soeder and @TheisMichaela - first big photo op then party with >2k people and Spider Murphy Gang 🕺🎉🥳

Great @nytimes read on #ChatGPT in education. ‘schools should thoughtfully embrace ChatGPT as a teaching aid — one that could unlock student creativity, offer personalized tutoring, and better prepare students to work alongside A.I.’. Same for academia. nytimes.com/2023/01/12/tec…

How can we compare disease dynamics in heterogeneous, typically short high-dimensional time series? Led by Amit Frishberg and Shai @shenorr, we developed TimeAx, which enables discovery of shared dynamics across patients via multiple trajectory alignment. biorxiv.org/content/10.110…

What do you hope for in single cell biology in 2023?

Lovely: science power move 🤩 @xkcdComic xkcd.com/2713/

Interesting viewpoint on AI-based code generation and the ‚End of Programming‘ by @mdwelsh @CACMmag: ‚I think CS as a field is in for a pretty major upheaval few of us are really prepared for.‘ dl.acm.org/doi/pdf/10.114…

How are β-cells changing during diabetes and how different are common preclinical mouse diabetes models? Led by @KHrovatin together w @LickertHeiko we built an integrated scRNA-seq islet atlas across 9 datasets, allowing comparison of models to human. biorxiv.org/content/10.110…

Beautiful perspective @NatureMedicine about the impact of the Human Cell Atlas on medicine by @jennyErood with Sarah @teichlab and Aviv Regev. Great summary of resources as well as open challenges! nature.com/articles/s4159…

The Institute for Stroke & Dementia Research @LMU_Muenchen together with us @CompHealthMuc is looking for a Junior PI in „Data Science for computational genomics, single-cell and spatial omics“ - fun intersection of bio/compute! Please consider applying! nature.com/naturecareers/…

Happy to announce ai.bay 2023, our Bavarian International Conference on AI, taking place at @DeutschesMuseum Feb 23+24 next year. Fun event with great science, speakers, exhibitions & politics - please consider joining/submitting abstract. #aibay2023 aibay2023.de

Been wondering what a good notion of a „spatial organ atlas“ could be: just a collection of microscopy+omics slices, for single or multiple samples? imputed cell types + annotation? identified niches? or patterns shared across samples akin to our cell type metaphor in scRNA-seq?

Nice 10-tips-article by @Benjamin_D_Lee et al @PLOSCompBiol on deep learning in biology. I like the opening of the box focus to „maximize interpretability“ - ideally we not only predict but want to learn about biology from our models. journals.plos.org/ploscompbiol/a…

How much can spatial information help to predict (disease) phenotypes? Led by @davidsebfischer & @mayarbali w @erturklab, we show that graph neural networks can learn emergent tissue properties from spatial molecular profiles. biorxiv.org/content/10.110… github.com/theislab/tissue

Deeply honored & excited to receive @dfg_public’s #Leibniz award 2023! I am continually amazed by the incredible individuals I have the privilege of working with, from my lab to colleagues @HelmholtzMunich, @TU_Muenchen & beyond. Big thanks to them, and most of all to my family!