Sign in

Moritz Schaefer

@moritzbaio.bsky.social
83 followers 119 following 31 posts

My Bio? AI Bio! Currently in the Bock lab@CeMM talking to single cells with LLMs. Here: paper reviews in 3 lines Mastodon (inactive): @moritzschaefer@qoto.org Twitter: @muronglizi

PostsRepliesMedia
Moritz Schaefer @moritzbaio.bsky.social · 08/03/2026
Want to analyze your single-cell data, but don't like to deal with code? 🧬 Good news: CellWhisperer now runs on your MacBook! 💻 Get your single-cell AI assistant running in under 5 minutes. ⏱️ Made with ❤️ @bocklab.bsky.social Code and docs at cellwhisperer.bocklab.org
010
Moritz Schaefer @moritzbaio.bsky.social · 11/11/2025
📝Finally out: Chat with your cells in English language - right in the browser. ✨ Try it yourself: cellwhisperer.bocklab.org and check out our tweetorial. Huge shout-out to to all the other contributors for the amazing teamwork!
cellwhisperer.bocklab.org
CellWhisperer: Unlocking the Language of Cells
030
Moritz Schaefer @moritzbaio.bsky.social · 21/07/2025
Just presented our new multimodal histopathology method "SpotWhisperer" at ICML, one of the largest AI conference. SpotWhisperer enables spatially resolved annotation of histopathology images using natural language. We achieved this by "transferring" annotations from transcriptomic data. More soon!
052
Moritz Schaefer @moritzbaio.bsky.social · 18/06/2025
After an exciting 3.5 years at CeMM and the AI Institute, it's time for me to move on to the next adventure. I am grateful for the friendships I made and the science I got to contribute to. I also wrote a brief recap of my science and activities during that time. Enjoy! moritzs.de/cemm
moritzs.de
Moritz Schaefer - BioMedical AI
Moritz Schaefer - BioMedical AI
000
Moritz Schaefer @moritzbaio.bsky.social · 06/05/2025
Having fun at work... :) Made possible by an amazing bunch of colleagues at my institute.
020
Moritz Schaefer @moritzbaio.bsky.social · 25/03/2025
Great behind-the-scenes overview on 'Flow matching and OT in cell biology' hrovatin.github.io/posts/sc_flo...
hrovatin.github.io
Matching flows to data
Models that fit like water in a vessel.
010
Reposted by Moritz Schaefer
Scientists4Future Österreich @s4fat.bsky.social · 07/03/2025
THANK YOU to everyone who joined the #March for #Science in #Vienna today! 💐💐💐 >500 people in front of @univie.ac.at, ~1000 people at Votivpark! 💚💚💚 Science is for everyone! #standupforscience2025 #scienceforall #sciencenotsilence
Crowd of protestors at March for Science, Vienna
13516
Reposted by Moritz Schaefer
Stand Up for Science! @standupforscience.net · 06/03/2025
SPEAKER LINEUP FOR STAND UP FOR SCIENCE ☀️ Here's our list of current speakers—and it's still growing! Come tomorrow to hear some amazing speakers talk about science, their experiences, and the future. #standupforscience2025
Speaker lineup for stand up for science - 


Bill Nye, Dr. Atul Gawande, Dr. Phil Plait, Dr. Michael Mann, Dr. Itai Yanai
Dr. Francis Collins, Dr. Victor Ambros, Dr. Nate Brought, Bill Nelson, JD
Dr. Ticora V Jones, Dr. Gretchen Goldman, Dr. Haley Chatelaine
Dr. Bonnielinn Swenor, Dr. Allison Agwu, Dr. Patrick Sullivan, Andrew Eneim
Dr. Nathaniel Kupperman, Bridget Nelson, MPH, Samantha Jade Duran
Denali Kincaid, Connor Phillips, Gabriella Amaya, Emily Whitehead
Sarah Aleman
Senator  Chris Van Hollen (MD), Rep. Bill Foster (IL-11) 
Rep. Suhas Subramanyam (VA-10), Rep. Jamie Raskin (MD-8)
 Rep. Glenn Ivey (MD-4), Former Rep. Fred Upton (MI-6)
4567150
Moritz Schaefer @moritzbaio.bsky.social · 16/01/2025
Yes. LLMs should rather be seen as "interface modules" of more complex AI systems: Their core role will be to *translate* between AI language (bits and bytes) and human-readable language.
000
Moritz Schaefer @moritzbaio.bsky.social · 30/12/2024
Visiting the #38C3 conference was an absolute blast again! Four days of great interactions and talks around politics, society, information security and science - My highlights post on LinkedIn: www.linkedin.com/posts/moritz... - I gave a talk too: lnkd.in/dXRF2mai - All talks here: media.ccc.de
lnkd.in
LinkedIn
This link will take you to a page that’s not on LinkedIn
000
Reposted by Moritz Schaefer
Peter Koo @pkoo562.bsky.social · 23/12/2024
I'm hiring: 1. Research associate (wet-lab w/ phd) to generate mpra perturbation data 2. ML postdoc to build multimodal generative AI for DNA (eg diffusion and LLMs) 3. Bioinformatician (any level) to process and harmonize functional genomics data to train foundation models DM me if interested!
12416
Reposted by Moritz Schaefer
Peter Gleick @petergleick.bsky.social · 25/12/2024
If you've never donated to Wikipedia, now's the time. Elon Musk is attacking it because it's one of the only sources of honest, curated information that he can't control. www.newsweek.com/elon-musk-ta...
newsweek.com
Elon Musk takes aim at Wikipedia
The billionaire has become one of the most prominent supporters of President-elect Donald Trump.
421122196347
Moritz Schaefer @moritzbaio.bsky.social · 19/12/2024
Thrilled to see our work with the @bocklab.bsky.social featured in 'The Scientist'! It's a really great read and builds on an interview I had with the @kamalnahas.bsky.social. www.the-scientist.com/an-ai-lab-pa...
the-scientist.com
An AI Lab Partner Helps Sift Through Transcriptomics Data
Big omics datasets can be overwhelming for researchers with limited programming skills, but texting with a new AI chatbot could help them wade through their results.
163
Moritz Schaefer @moritzbaio.bsky.social · 21/11/2024
Had a fantastic time last week at CSHL's #biodata24, with amazing discussions around AI and bioinformatics and presenting my work on CellWhisperer. Thanks @mike_schatz and all organizers, looking forward to being back in 2026!
010
Moritz Schaefer @moritzbaio.bsky.social · 19/11/2024
How come everyone is joining Bluesky these days? I thought it dead already.
000
Reposted by Moritz Schaefer
Stephanie Hicks @stephaniehicks.bsky.social · 15/11/2024
Congratulations to @findycang.bsky.social from @jhubiostat.bsky.social who did an amazing job presenting her work at #biodata24 tonight! Preprint coming soon. ☺️ #ProudPI #stats #spatial #omics #transferlearning 🧪🧬🖥️🧠📈
Cindy Fang presents her work at CSHL Biological Data Science
0484
Reposted by Moritz Schaefer
Andre Rendeiro @andre-rendeiro.com · 16/11/2024
What if we could look at a human tissue image🔬and quantify its biological age?🧬 And how could we get the same insights without collecting the tissue sample? In our new @biorxivpreprint, we create "tissue clocks" ⏳to do just that! www.biorxiv.org/content/10.1... 📝 #aging #tissue thread below⬇️
23713
Moritz Schaefer @moritzbaio.bsky.social · 15/08/2024
"Contextual AI models for single-cell protein biology" 🧪🧬🖥️ - Protein embeddings are powerful. - But: they are usually cell-type agnostic - This paper: Embedding cell-type-aware protein representations via GNNs on scRNA-seq-pruned protein interaction graphs www.nature.com/articles/s41...
nature.com
Contextual AI models for single-cell protein biology - Nature Methods
PINNACLE is a context-specific geometric deep learning model for generating protein representations. Leveraging single-cell transcriptomics combined with networks of protein–protein interactions, cell...
000
Moritz Schaefer @moritzbaio.bsky.social · 13/07/2024
"Transformers need glasses! Information over-squashing in language tasks" - In typical LLMs every token only 'sees' previous tokens (causal attention) - This leads to information inbalance across tokens - Thereby, later tokens 'vanish' in long sequences arxiv.org/pdf/2406.04267
000
Moritz Schaefer @moritzbaio.bsky.social · 25/06/2024
'The primitive endoderm (PrE) supports lineage plasticity to enable regulative development' 🧬️🧪 - PrE is extra-embryonic tissue (at day 4) but can still recreate a full blastocyst (i.e. also embryo) - JAK/STAT supports this plasticity - www.cell.com/cell/abstrac...
cell.com
The primitive endoderm supports lineage plasticity to enable regulative development
Although generally considered a simple support tissue, extra-embryonic primitive endoderm has the capacity to regenerate a complete blastocyst and continue post-implantation development.
000
Moritz Schaefer @moritzbaio.bsky.social · 10/06/2024
"Why is exercise good for you?" - Humans, as hunter-gatherers, evolved to 'always move' - Exercise alters gene expression throughout the body - With exercise, muscles contribute to anti-inflammation through IL-6 production www.nature.com/articles/d41... 🧬🖥️🧪
nature.com
Why is exercise good for you? Scientists are finding answers in our cells
Decades of evidence shows that exercise leads to healthier, longer lives. Researchers are just starting to work out what it does to cells to reap this reward. Decades of evidence shows that exercise l...
020
Reposted by Moritz Schaefer
Richard M. Carpiano, PhD, MPH @rmcarpiano.medsky.social · 07/06/2024
Cool study... Post-January 6th deplatforming on Twitter reduced the reach of misinformation Hats off to my UCR colleagues Kevin Esterling and Diogo Ferrari and their co-authors... 🧪sociology polisky policysky www.nature.com/articles/s41...
nature.com
Post-January 6th deplatforming reduced the reach of misinformation on Twitter - Nature
Difference-in-differences analysis indicates that the decision by Twitter to deplatform 70,000 users following the events at the US Capitol on 6 January 2021 had wider effects on the spread of mi...
04721
Moritz Schaefer @moritzbaio.bsky.social · 05/05/2024
openreview.net/pdf?id=KIeU4... - Most drug design methods assume rigid proteins - Simple docking experiments show: Flexible side chains enhance modeling - New method FlexFlow: Flow matching of small molecules in addition to side chain configurations Looking forward to full release 🧬 & 🖥️
openreview.net
000
Moritz Schaefer @moritzbaio.bsky.social · 03/05/2024
I go 90% with the stated opinion about OpenAI etc. Yet, it's a real pity that some arguments needed to be overdrawn. I use GPT-4 daily and find it extremely useful. Also, highly unlikely that LLMs will worsen due to data pollution. One more example of how polarisation wins the attention market..
010
Moritz Schaefer @moritzbaio.bsky.social · 10/04/2024
- A common problem in ML is to generalize to cases unseen during training - Here, GPT- and protein embeddings enable perturbation effect prediction for unseen genes - My take: Great direction and I anticipate more 'embedding arithmetics' (e.g. 'gene1-gene2' for overexpression and KO)
openreview.net
Enhancing generative perturbation models with LLM-informed gene...
Genetic perturbations are key to understanding how genes regulate cell behavior, yet the ability to predict responses to these perturbations remains a significant challenge. While numerous...
010
Moritz Schaefer @moritzbaio.bsky.social · 08/04/2024
The original 'oversquashing' paper is very accessibly written: - In GNNs, when data needs to pass many nodes, the data gets exponentially 'diluted' - Adding a simply fully-connected edge layer, improved upon QM9 by 42% - This foundational and simple concept is only from 3 years ago
arxiv.org
On the Bottleneck of Graph Neural Networks and its Practical Implications
Since the proposal of the graph neural network (GNN) by Gori et al. (2005) and Scarselli et al. (2008), one of the major problems in training GNNs was their struggle to propagate information...
010
Moritz Schaefer @moritzbaio.bsky.social · 04/04/2024
"Cell subtype-specific effects of genetic variation in the Alzheimer’s disease brain" - scRNA-seq 🧬 allows high-throughput variant effect description 🖥️ - Here: brain tissue from 424 individuals, revealing unprecedented number of eQTLs - Analysis mostly promoter-driven. More Hi-C data needed?
nature.com
Cell subtype-specific effects of genetic variation in the Alzheimer’s disease brain - Nature Genetics
Single-nucleus RNA sequencing from the dorsolateral prefrontal cortex of 424 aging individuals, and mapping the effect of genetic variation on gene expression, identified a large number of cis-express...
020
Moritz Schaefer @moritzbaio.bsky.social · 03/04/2024
"Understanding Biology in the Age of Artificial Intelligence" arxiv.org/abs/2403.04106 AI models are uncommon to uncover strict "laws" in biology. How are they still useful? An epistemological perspective 3-bullet-point review in the comment below
110
Moritz Schaefer @moritzbaio.bsky.social · 01/04/2024
cos_sim(a, b) = dot(a_norm, b_norm) But is it always better for comparing embeddings? - No: cos_sim may yield arbitrary results - Cause: Model expression flexibility - Remedy: Train with cosine similarity, or normalize dataset *before* embedding learning
arxiv.org
Is Cosine-Similarity of Embeddings Really About Similarity?
Cosine-similarity is the cosine of the angle between two vectors, or equivalently the dot product between their normalizations. A popular application is to quantify semantic similarity between...
000
Moritz Schaefer @moritzbaio.bsky.social · 31/03/2024
Apple's insights to training multimodal (image) LLMs: - LLaVA-like approach (fine-tune LLM on tokenized images) - Image encoder model is key. More information (image feature dimensions and tokens) = better - Augment text-only or image-interleaved documents to prevent deteriorating LLM performance
arxiv.org
MM1: Methods, Analysis & Insights from Multimodal LLM Pre-training
In this work, we discuss building performant Multimodal Large Language Models (MLLMs). In particular, we study the importance of various architecture components and data choices. Through careful...
000
Moritz Schaefer @moritzbaio.bsky.social · 29/03/2024
Geometrical explanation of LayerNorm [(x-μ)/σ] arxiv.org/abs/2305.025... - implicitly projects inputs orthogonal to 1-vector and scales to norm of sqrt(d) - enables transformers to attend to all keys simultaneously (sometimes useful) - ensures that attention keys are not too "small" to be selectable
010
Moritz Schaefer @moritzbaio.bsky.social · 28/03/2024
www.biorxiv.org/content/10.1... - Aggregate scRNA-seq dataset (human and mouse) with annotated cell potency (i.e. totipotent, pluripotent, multipotent, unipotent). - Train a neural network to predict level of potency. - Neural network has binary weights, allowing to read out predictive genes
020