Mafalda has been shortlisted for the 2026 Premio Vanguardia de la Ciencia, one of the more visible science prizes in Spain. Her interview: www.lavanguardia.com/vanguardia-d... ...and we would be very grateful if you could vote here: www.lavanguardia.com/vanguardia-d...

How much does a language model forget when finetuned on new tasks? We show both model size and optimization matter and forgetting can be nearly eliminated with self-generated replay! arxiv.org/abs/2605.26097 w/Martin Marek, Dongkyu Cho, Shikai Qiu, Rumi Chunara, and Pavel Izmailov. 1/8

AI drug discovery has a data problem. We're fixing it. At ALLOX, we've spent the last few years building something that didn't exist before: a platform that systematically maps functional hotspots at scale, measuring the effects of mutations on protein-protein interactions across the proteome.

Join us for our next Kipoi Seminar with Noor Youssef, Harvard Medical School @harvardmed.bsky.social @deboramarks.bsky.social 👉 Predicting Viral Evolution to Accelerate Vaccine and Therapeutic Design 🧬 Wed May 6, 5:30pm CET 🍃http://kipoi.org/seminar 🦋 @kipoizoo.bsky.social

Very nice paper from Regeneron www.nature.com/articles/s41...

🧬 What if reading a tumor's mutations could tell us exactly how to fight it? Dr. @nlbigas.bsky.social is doing exactly that — decoding the genomic fingerprints of #cancer to guide personalized treatments. This is #CancerResearch, reimagined.

Meet evedesign: open-source AI, accessible protein design ✅Combine models for multiobjective optimization ✅Integrate experimental data ✅ Run on your own infrastructure 📄Paper: www.biorxiv.org/content/10.6... 💻Code: github.com/evedesignbio 🌐Webserver: evedesign.bio Collaborate: [email protected]

New paper showing that much of the apparent success of protein language models in predicting mutational effects is a mirage: These models mostly memorize sites. 1/ www.biorxiv.org/content/10.6...

The main project of my PhD 🧬🔬 is out: we developed single-cell lentiMPRA, a lentivirus-based method to measure enhancer activity and transcriptomes at single-cell resolution. We then applied sc-lentiMPRA to fully synthetic enhancers 🧩... 🔗 doi.org/10.64898/202...

🚨 NEW @FraticelliLab PREPRINT JUST WENT LIVE biorxiv.org/content/10.6... Here, we describe PLSTCs, Soluplus-based expansion cultures of primary mouse *Acute Myeloid Leukemia* Stem Cells. And then we use LARRY barcoding to identify drivers of leukemogenicity and drug resistance!

Join us at Cell Symposia: Single-Cell Biology in the Era of AI (Dec 2–4, 2026, Munich) 🇩🇪 From multi-omics to AI-driven models of cellular dynamics — an exciting lineup across experimental & computational biology. Hope to see you there! 🔗 www.cell-symposia.com/single-cell-...

🚀 From modality-specific to compositional foundation models for cell biology Happy to share our new Perspective in Cell Systems. We describe an emerging shift toward compositional multimodal architectures 🧩 authors.elsevier.com/c/1mdzg8YyDf...

Early bird registration is closing Feb 28! We will have an AMAZING party!

Our grad-level "Deep Learning" course (MIT's 6.7960) is now freely available online through OpenCourseWare: ocw.mit.edu/courses/6-79... Lecture videos, psets, and readings are all provided. Had a lot of fun teaching this with @sarameghanbeery.bsky.social and @jeremybernste.in!

Post 3 of attempting to share more ML/Bio papers on Bluesky: Accelerating Scientific Research with Gemini: Case Studies and Common Techniques arxiv.org/abs/2602.03837

Training LLMs with verifiable rewards uses 1bit signal per generated response. This hides why the model failed. Today, we introduce a simple algorithm that enables the model to learn from any rich feedback! And then turns it into dense supervision. (1/n)

Post 2 of sharing interesting ML/bio papers that don’t seem to be getting much attention on Bluesky. Three papers on self-distillation, all from the same week! arxiv.org/abs/2601.18734 arxiv.org/abs/2601.19897 arxiv.org/abs/2601.20802

Why don’t neural networks learn all at once, but instead progress from simple to complex solutions? And what does “simple” even mean across different neural network architectures? Sharing our new paper @iclr_conf led by Yedi Zhang with Peter Latham arxiv.org/abs/2512.20607

Bluesky doesn’t really shine yet as a place to discover papers, especially at the intersection of biology and ML. To help a little, I’m going to start sharing papers I find interesting. To kick things off, here’s one from @petar-v.bsky.social and colleagues at DeepMind arxiv.org/abs/2601.22950

"Probing the Dark Energy in the Functional Protein Universe" is now published in @pnas.org 🎉 www.pnas.org/doi/10.1073/...

For no reason, I remembered today that I too once got to take a picture holding Nobel Prize that I didn't earn

TF-MINDI is out! A new method to learn cis-regulatory codes through rich embeddings of TF binding sites. TF-MINDI decomposes motif neighbourhoods, and works downstream of any sequence-to-function deep learning model. We deeply study the enhancer code in human neural development, check out the thread

We’ve updated the EVEREST benchmark to include real-world viral evolution! www.biorxiv.org/content/10.1... Co-led by Noor Youssef and me, along with co-authors Navami Jain, Aarushi Mehrotra, Sarrah Leung, Abigail Jackson, @deboramarks.bsky.social, and with @cepi.net @futurehousesf.bsky.social!

🚨🚨🚨 We're looking for a research tech to work on alt splicing, pancreatic islets and diabetes. The goal is to set a high-throughput platform to investigate the role of alternative exons in beta cell biology! Interested in joining our lab at @melisupf.bsky.social? 👇 www.upf.edu/documents/d/...

Excited to share a preprint of my PhD project looking at interactions between SNPs and polygenic scores in the UK Biobank! A thread... 🧵 www.medrxiv.org/content/10.1...

LFB is NeurIPS-bound! 🎉 Mafalda, @cwjpugh.bsky.social and I will be in San Diego next week for NeurIPS -- happy to chat variant effect prediction (or just say hi). “From Likelihood to Fitness: Improving Variant Effect Prediction in Protein and Genome Language Models” openreview.net/pdf/a151f62e...

New paper “Proteome-wide model for human disease genetics” is now live at Nature Genetics: rdcu.be/eRu7K popEVE (pop.evemodel.org) finds the needles in the haystacks of human genetic variation:

popEVE is out in Nature Genetics! 🎉 We built a proteome-wide model that combines cross-species and human population variation to rank missense variants by disease severity and help diagnose rare genetic disorders. rdcu.be/eRu7K

PopEVE covered by @financialtimes.com today! www.ft.com/content/bc49...

Today we introduce Mafalda Dias and @jonnyfrazer.bsky.social from CentreGenomic Regulation @crg.eu ,with one PhD position open in their lab. Learn more about the PhD programme and how to apply → www.evomg-dn.eu #PhD #DoctoralTraining #ResearchCareers #LifeSciences #Genomics #EvolutionaryBiology

🚨 Two weeks left to apply!

🚨 15 days left to apply! Be part of EvoMG-DN and advance your PhD career in evolutionary genomics. 👉 www.evomg-dn.eu #PhD #DoctoralTraining #ResearchCareers #LifeSciences #BiomedicalResearch #Genomics #EvolutionaryBiology #MSCA #HorizonEurope #EUResearch #ResearchOpportunities #PhDPositions

🚀 Excited to share our preprint scConcept: a contrastive pretraining framework for technology-agnostic single-cell representations. We move beyond reconstruction to learn rich, cell-level embeddings for diverse omics. 📘 www.biorxiv.org/content/10.1...

First time on Bsky and first big announcement! I am excited to announce that our new study explaining the missing heritability of many phenotypes using WGS data from ~347,000 UK Biobank participants has just been published in @Nature. Our manuscript is here: www.nature.com/articles/s41....

(1/N) Thrilled to share that our paper HiPoNet (High dimensional Point cloud Network) to be presented at NeurIPS 2025! HiPoNet treats an entire high-dimensional point cloud as a datapoint! It captures multi-scale geometry and topology of the cloud perform classification and regression tasks.

More than two decades have passed since we discovered that rare disruptions of the FOXP2 gene disturb development of proficient speech/language skills. Today we know of multiple FOXP genes that are directly implicated in distinct brain-related conditions with differences in symptoms & severity.🧪 1/n

Our review is out in Nature Reviews Genetics! rdcu.be/d5AY2 We show how phylogeny-based methods can resolve the problem of non-independence in genomic datasets. These methods must be considered an essential part of the comparative genomics toolkit. @lauriebelch.bsky.social @stuwest.bsky.social

Now online @nature.com! Want to change the consequences of receptor activation? Small molecules binding the GPCR-transducer interface change G protein subtype preference in predictable ways, enabling rational drug design 💥 So many new possibilities! 🧪🧠🟦 www.nature.com/articles/s41... 🧵👇

Applications are open for the @crg_eu PhD Programme! 20 fully funded positions — including one in our group through the Evolutionary Medical Genomics ITN. Join us to develop deep generative models of cross-species data to tackle open questions in disease genetics. www.crg.eu/en/content/t...

Announcing our new protein design server evedesign.bio: • End-to-end protein design for everyone! • Analyze your generated library interactively and on 3D structures • Export codon-optimized DNA sequences for experimental testing.

The second session (Thr 09/10, morning) was dedicated to the Genomic Basis & History of Disease.

I adored writing this piece. It brings together several of the things preoccupying me right now, like chromatin organization and gene regulation. There's so much more to be said on that. Also, these marine critters look gorgeous. www.quantamagazine.org/loops-of-dna...

The most beautiful data is on the horizon!

This is truly an incredible breakthrough IMO. Really exemplifies what you get when deep domain expertise (popgen/evolution/disease genetics in this case) fuses with cleverly crafted ML. What u get r sleek, well thought out architectures that absolutely destroy the behemoths. Wow!! 1/

Most of the #EurIPS workshops now have their websites online 🙌 A lot of them have active call for participation and/or papers, so if you see something relevant to your field of research please consider submitting! Links for most workshops are available at: eurips.cc/workshops/

We are excited to share GPN-Star, a cost-effective, biologically grounded genomic language modeling framework that achieves state-of-the-art performance across a wide range of variant effect prediction tasks relevant to human genetics. www.biorxiv.org/content/10.1... (1/n)

Excited for a major milestone in our efforts to map enhancers and interpret variants in the human genome: The E2G Portal! e2g.stanford.edu This collates our predictions of enhancer-gene regulatory interactions across >1,600 cell types and tissues. Uses cases 👇 1/

NeurIPS 2025 financial assistance and volunteer applications are now open for both San Diego and Mexico City! The deadline for both applications is 1 Oct AoE. See neurips.cc/Conferences/... for more details.

Many of the most complex and useful functions in biology emerge at the scale of whole genomes. Today, we share our preprint “Generative design of novel bacteriophages with genome language models”, where we validate the first, functional AI-generated genomes 🧵

Exciting to see our protein binder design pipeline BindCraft published in its final form in @Nature ! This has been an amazing collaborative effort with Lennart, Christian, @sokrypton.org, Bruno and many other amazing lab members and collaborators. www.nature.com/articles/s41...