Saez-Rodriguez Group

👋 @veronicalombardi.bsky.social has completed her 6-month visit period with us. She is a PhD student from the University of Rome La Sapienza, and her contribution focused on generating single-cell and patient-specific networks of pancreatic cancer using our tools LIANA+ and CORNETO. All the best! 🍀

Current Vacancies Whether you're a scientist, IT specialist, accountant or administrator, you'll help us tackle the challenges of improving human health & biodiversity in the face of climate change on a global scal...

We are hiring a Postdoctoral Fellow in Computational Biology at EMBL-EBI (Cambridge, UK). Focus: methods to study cell–cell communication from sc/spatial omics data (building on LIANA+ and NicheNet), in collab with @yvansaeys.bsky.social VIB/Ghent.

Details & apply by 13/10/25: tinyurl.com/4shdw8dk

It was also supported by projects: AI4FOOD-CM (Y2020/TCS6654), FACINGLCOVID-CM (PD2022-004-REACT-EU), INTER-ACTION (PID2021-126521OB-I00 MICINN/FEDER), HumanCAIC (TED2021-131787BI00 MICINN), PowerAI+ (SI4/PJI/2024-00062 Comunidad de Madrid and UAM), and Cátedra ENIA UAM-VERIDAS.

This work was supported through state funds approved by the State Parliament of Baden-Württemberg for the Innovation Campus Health + Life Science Alliance Heidelberg Mannheim.

This work was a nice collaboration with Ph.D. student Sergio Romero, supervised by professors Ruben Tolosana and Aythami Morales (UAM, Spain), who was visiting us for 3 months to work on this project with @pablormier.bsky.social and @martingarridorc.bsky.social.

GitHub - scapeML/scape: Single-cell Analysis of Perturbational Effects using Machine Learning Single-cell Analysis of Perturbational Effects using Machine Learning - scapeML/scape

🔧 We developed ScAPE using #Keras 3, so you can use #Tensorflow, #JAX or #PyTorch as backends
👉 github.com/scapeML/scape

We benchmarked ScAPE against:
🔹 Other winning challenge methods
🔹 TabPFN, a foundation model for tabular data

➡️ ScAPE matches or outperforms them, showing the value of simple, efficient baselines.

Despite its simplicity, ScAPE ranked among the top methods in the challenge.
It generalizes across new drug–cell combinations and offers a robust baseline for evaluating novel approaches.

✨ ScAPE (Single Cell Analysis of Perturbational Effects)

- Lightweight neural network (∼19M params)
- Uses only aggregated gene-level stats (robust + simple)
- Multi-task: predicts both significance (p-values) & effect size (fold-change)

Predicting how cells respond to drugs is central to drug discovery & precision medicine. But existing models often struggle to generalize, and don’t always beat simple baselines.

🚨 New preprint

We present an extended version of ScAPE, the method that won one of the prizes 🏆 in the @neuripsconf.bsky.social 2023 Single-Cell Perturbation Prediction challenge.

📄 preprint: doi.org/10.1101/2025...
🧬 code: github.com/scapeML/scape

We also gratefully acknowledge our funding support from:
• The Landesinstitut für Bioinformatikinfrastruktur in Baden-Württemberg
• The German Research Foundation (DFG)
• The European Union’s Horizon 2020 Programme
• UKRI Biotechnology and Biological Sciences Research Council (BBSRC)
• TUBITAK ARDEB

👏 A huge thank you to all 32 authors from @saezlab.bsky.social, @ebi.embl.org and Heidelberg University, Korcsmaros group #ICL, Tunca Dogan’s team at Hacettepe University, Michal Klein @fabiantheis.bsky.social, Francesco Ceccarelli, and everyone else for their fantastic work!

🎯 OmnipathR is more than a client—it's a 🛠️ Swiss Army knife for prior knowledge. Access OmniPath and 25+ more databases, translate IDs, and get dedicated support for #metabolomics
➡️ Check it out: r.omnipathdb.org #RStats

📊 OmniPath was built for #NetworkBiology. Use it for:
• Network analysis directly in Cytoscape (apps.cytoscape.org/apps/omnipath)
• Mechanistic modeling with Corneto (corneto.org) and NetworkCommons

🔬 Perfect for #singlecell and #spatialomics analysis! The 🐍 #Python client plugs directly into @scverse.bsky.social. Use OmniPath with tools like 🧩 decoupler.readthedocs.io and 🧩 liana-py.readthedocs.io for pathway activity prediction, automated cell annotation, cell-cell communication and more.

⚙️ Integrate @omnipath.bsky.social directly into your #Bioinformatics workflows. We provide a web service, and powerful clients for R (@bioconductor.bsky.social), Python (@scverse.bsky.social), and Cytoscape. License compliance for companies built-in! ⚖️ 🤑

🤖✨ Dive into OmniPath Explorer - Interactively browse millions of annotations & hundreds of thousands of molecular interactions. Our new AI chat agent helps you query the database effortlessly, developed by Jonathan @jonsch.one!

🧬 What's inside? Literature-curated, mechanistic data is our core:
• Signaling & gene regulatory pathways
• miRNA interactions & enzyme-PTM relationships
• Protein complexes & cell-cell communication
• Functional, localization, & structural annotations
...plus high-throughput & predicted knowledge.

🥁 Check out our new preprint on OmniPath, the prior knowledge resource for #SystemsBiology, and its brand-new OmniPath Explorer web app! 🥳

📖 Preprint: www.biorxiv.org/content/10.1...
🔍 Explorer: explore.omnipathdb.org

OmniPath integrates 160+ resources for multi-omics analysis & modeling.

🧶⬇️

This work was led by @psl-schaefer.bsky.social, with support from Leoni Zimmermann, Leo Burmedi, @tanevski.bsky.social ‬, Miri Adler and her team (A. Walfisch, N. Goldenberg, S. Yonassi, E. Shaer Tamar). @juliosaezrod.bsky.social ‪‬and @ricoramirez.bsky.social supervised the work.

Beyond single-cell measurements, ParTI has been applied to cognition, behavior, and morphology. With ParTIpy, these applications become more accessible through a modern, scalable Python implementation. Have a try and share with us your feedback.

scverse (@scverse.bsky.social) Foundational tools for omics data (mostly in python).

ParTIpy provides tools to assess model fit, explore archetypes, and run downstream analyses. Features like enrichment, cell–cell communication, and spatial mapping support single-cell data, while scverse.bsky.social integration ensures compatibility with other tools.

ParTIpy scales this framework to large-scale datasets by combining advances in initialization, optimization and coreset methods for archetypal analysis, the algorithmic backbone of ParTI.

Uri Alon (@urialonlab.bsky.social) The Alon lab explores design principles of biological circuits, focusing on systems medicine, aging, and healthspan.

Pareto Task Inference (ParTI), pioneered by @urialonlab.bsky.social team, models variability as trade-offs between archetypal functions. Unlike clustering, it captures continuous variation in a principled way. Offering an alternative to cell-state definitions, for example.

Introducing ParTIpy, a python package for Pareto Task Inference that scales to large-scale datasets, including single-cell and spatial transcriptomics.
🔗 Manuscript: www.biorxiv.org/content/10.1...
💻 Code: partipy.readthedocs.io

@embl.org PhD programme just open - many interesting projects, including two with us @saezlab.bsky.social @ebi.embl.org - one on #host-microbiome interactions using #spatial omics, and one on #AI in #proteomics 👇

👋 Time to say goodbye to another one of our PhD students, @smuellerdott.bsky.social. She was working on the integration of phosphoproteomics and transcriptomics data and will now join the Computational Oncology Team at Merck. We wish her all the best for this new chapter! 🚀 ✨

🙌 Thanks also to everyone who made their data publicly available and to our funders SmartCare (BMFTR), CRUK Programme Foundation, Alexander von Humboldt Foundation and the Landesinstitut für Bioinformatikinfrastruktur in Baden-Württemberg.

➡️ We welcome feedback, issues & contributions!

🙌 We also thank Robin Fallegger, Charlotte Boys, Christoph Mahler, Aurelien Dugourd, Yunfan Bai and Jannik Franken for valuable feedback.