We’re so excited to share our new paper, where we tackle the wealth of structural and functional diversity of anti-phage sensors across bacteria (1/6)

Nature@nature.com · last wk.

Nature research paper: Diverse bacterial pattern recognition receptors sense the core phage proteome go.nature.com/45fpXCZ

🚨 JOB ALERT🚨 We are very excited to be hiring the lab's ✨very first postdoc✨! Work on new AI technologies for decoding antigen protein evolution in a fresh research environment, at the heart of Paris 🇫🇷 Details & link to apply: research.pasteur.fr/en/job/postd... Deadline: Sep 1st

New review out! At @annualreviews.bsky.social for Virology! Check out the summary thread below 👇👇👇

Antigen Evolution & Design Lab@antigenevo.bsky.social · 3w ago

First paper with the lab's affiliation!! ✨✨ summarising some of the cool concepts that the Antigen Evolution & Design Lab will be working on! Together with amazing collaborators @mghafari.bsky.social and @grovearmada.bsky.social www.annualreviews.org/content/jour...

We often talk about complex protein folds but what does that really mean? Turns out we don't have a good metric for it. So we made a game to make one! Head over to our website to vote on which fold you think is more complex and why! pacesalab.com/foldcomplexi...

Inaugural post! We are a brand new lab @pasteur.fr studying how pathogen antigens evolve and how we can exploit these proteins' evolution to make better vaccines! Read a few words from the currently sole member of the lab @spyroslytras.bsky.social below! 😁 www.pasteur.fr/en/research-...

Predicting viral evolution with AI

Spyros Lytras is one of four young scientists selected in 2025 to lead new research groups at the Institut Pasteur. He has headed the “Antigen Evolution & Design laboratory” 5-year Group (G5) since Ju...

pasteur.fr

I gave a workshop on AlphaFold & related programs @ the University of Utah. Part 1 covers protein structure prediction before/after AlphaFold & deep learning neural networks. Part 2 covers the math behind the scores. Part 3 covers PyMOL & the webservers. youtube.com/@rolanddunbr...

New Experimental Google Colab Notebook now integrates AlphaFold3 with OpenFold3 and py2Dmol: colab.research.google.com/github/sokry... (1/3)

Folddisco is now published @natbiotech.nature.com. It’s a fast motif search for similar 3D DISCOntinuous residues like catalytic sites or zinc fingers across the entire protein universe. 📄 www.nature.com/articles/s41... 💾 folddisco.foldseek.com​​​​​​​​​​​​​​​​ 🌐 https://search.foldseek.com/folddisco

Structural motif search across the protein universe with Folddisco - Nature Biotechnology

Folddisco enables protein structural motif search in million scale databases.

nature.com

Martin Steinegger 🇺🇦@martinsteinegger.bsky.social · last yr.

Folddisco finds similar (dis)continuous 3D motifs in large protein structure databases. Its efficient index enables fast uncharacterized active site annotation, protein conformational state analysis and PPI interface comparison. 1/9🧶🧬 📄 www.biorxiv.org/content/10.1... 🌐 search.foldseek.com/folddisco

ESMFold2 👀

Nature@nature.com · 2mo ago

A newly released AI tool has generated an atlas of more than one billion predicted protein structures and billions more protein sequences. go.nature.com/4wS3Pen

🚀 Excited to share our new work: Absolute Stability Predictor! 📊: forms.gle/4ZnXZSnTBvay... Built the MGnify Stability Dataset (1.8M+ measurements) and developed stability prediction models, together with @grocklin.bsky.social @KotaroTsuboyama, @sokrypton.org and teams.

Congrats to @ulad-litvin.bsky.social on winning 2nd best talk at #Vibiom2026! @cvrinfo.bsky.social 🏅

Fifth and final @cvrinfo.bsky.social talk at ViBioM 2026. @ulad-litvin.bsky.social presenting viro3d.cvr.gla.ac.uk and other applications of alphafold in virology.

New work: Our earlier work showed that Foldseek characters could be adapted for phylogenetic alignment, treating each character as an evolutionary state. That part holds but there's a hidden assumption baked in that needed unpacking. So lets do that.🧵 #StructuralPhylogenetics #Evolution #Protein

With Eugene Koonin, we wrote a rather comprehensive review on the origin, evolution and organization of the #virosphere. We describe all 10 viral realms and the logic behind them, and so much more. Check it out! comptes-rendus.academie-sciences.fr/biologies/ar...

Had a great time at the Microbiology Society Annual Conference! I gave a talk on modelling protein-protein interactions with AlphaFold3. A few lessons learned 🧬🧵

Some bat alphacoronaviruses can exploit an entirely different receptor to enter human cells. This could lead to crossover events could be the first step in future epidemics, though we are not there yet and there would be many more barriers to overcome also. www.nature.com/articles/s41...

Heart-nosed bat alphacoronaviruses use human CEACAM6 to enter cells - Nature

Human CEACAM6, which is widely expressed in the lung, is identified as a receptor used by the spike proteins of Cardioderma cor (heart-nosed bat) alphacoronaviruses to enter cells.

nature.com

🎉Excited to share our new preprint! TIR domains from diverse animals, including human TLR4, are catalytically active and produce cyclic ADP-ribose (cADPR). This enzymatic activity is widespread in animals and conserved across the tree of life. www.biorxiv.org/content/10.6...

We introduce ConforNets, a mechanism for conformational control in AlphaFold3 models - SoTA at producing diverse conformations on every multistate benchmark (N=104) - Novel capability: transfer state from one protein to another Outperforms BioEmu, ConforMix and AFsample3 🧵1/8

"The selfish ribosome" paper is now published in @plosbiology.org . Have a look! Many thanks to the editors and reviewers! doi.org/10.1371/jour...

The selfish ribosome

In this Essay, the evolution of life is construed as a ribosomal takeover, whereby the ribosome evolved to consume most of the cell’s resources, while other cellular componentry ensured the propagatio...

journals.plos.org

Mart Krupovic@mkrupovic.bsky.social · 5mo ago

With Eugene Koonin, we propose a concept of “the selfish ribosome”, under which evolution of life is viewed as a ribosomal takeover, where the ribosome evolved to consume most of the cell’s resources, while other cellular componentry ensures the propagation of the ribosome. arxiv.org/abs/2602.23268

The CASP experiment is about to start - and is still short on challenging prediction targets. Ligand complexes, RNA, …

Torsten Schwede@torstenschwede.bsky.social · 5mo ago

Is #AI hitting a plateau in structure prediction? Help us find out at CASP17! 🧪🧬 Calling for Targets: Immune Complexes, protein - ligand complexes, RNA/DNA, conformational ensembles, membrane proteins, viral origins, and large complexes. The Rule of Thumb: If AF3 can’t model it, we want it.

Having fantastic time at @microbiologysociety.org annual conference in Belfast. Just gave a short talk on predicting protein-protein interactions with AlphaFold3 🔮 and all the tricks we’ve applied to make it work for challenging targets 👾

Joseph Hughes@bljog.bsky.social · 4mo ago

Great presentation from @ulad-litvin.bsky.social @microbiologysociety.org annual conference: How far can AlphaFold go? Lessons from modelling virus–host protein interactions using AlphaFold3.

45 novel protein folds in the updated AFESM (AFDB + ESMatlas) manuscript: • 12 high-confidence folds in AFESM • 33 by ColabFold-repredicting 2.3M low-quality domains We show AFDB captures most domains already and ESMfold struggles with novelty 🌏 afesm.foldseek.com 📄 biorxiv.org/content/10.1...

📣Huge preprint 🔔 Today we share something our group has been working toward for a long time, led by @lucasmoriniere.bsky.social We asked can we predict which receptor a phage targets from its genome sequence alone? For most phages, we couldn’t. So Lucas set out to do something I had only dreamed of.

How diverse is bacterial immunity ? We report in @science.org how language models allowed us to predict 2.4M antiphage proteins spanning >23K novel potential systems. 👏 @emordret.bsky.social, @alexhv.bsky.social & al doi.org/10.1126/scie... Explore them here defensefinder.mdmlab.fr/wiki/refseq_...

How can we study the function of nucleotide-based #SecondMessengers? @oliveiramann.bsky.social presented her group's exciting findings in this field at #Biochemistry2026. She e.g. presented the #cryoEM structure of OAS2, which has key #antiviral functions. linkinghub.elsevier.com/retrieve/pii...

Kieran et al. trained a generative protein binder design model. The training is based on Teddymer, a dataset developed by @sooyoung-cha.bsky.social. By treating monomer domains as multimers and clustering them with Foldseek, she created a set that allowed Complexa to learn. 💾 teddymer.foldseek.com

Kieran Didi@kdidi.bsky.social · 5mo ago

📢 We’re launching Proteina-Complexa — and after the Jensen keynote mention, we definitely had to post this thread now ;) Atomistic binder design with generative pretraining + test-time compute, plus large-scale wet-lab validation. Project page: research.nvidia.com/labs/genair/... 🧵 1/n

You asked, we listened. Millions of AI-predicted protein complex structures are now available in the #AlphaFold Database. This spans homodimers from 20 of the most studied species, including humans, as well as the World Health Organization’s priority pathogens list. www.ebi.ac.uk/about/news/t...

AlphaFold database has entered the era of complexes. Together with NVIDIA, DeepMind and EBI, we use ColabFold, OpenFold and MMseqs2-GPU to predict ~31 million complexes (homo & hetro-dimers) resulting in 1.8 million high-quality predictions 📄 research.nvidia.com/labs/dbr/ass... 🌐 alphafold.ebi.ac.uk

Having a hard time keeping up with the fast-moving field of origin of eukaryotes/Asgards archaea? Our new perspective article can help, we review the status of research over the last decade and where it's going. The archaeal roots of eukaryotic life www.pnas.org/doi/10.1073/...

New OpenFold3 preview out! (OF3p2) It closes the gap to AlphaFold3 for most modalities. Most critically, we're releasing everything, including training sets & configs, making OF3p2 the only current AF3-based model that is functionally trainable & reproducible from scratch🧵1/9