
The EMBL‑EMBO AI and Biology symposium in March 2026 highlighted a paradigm shift: artificial intelligence is moving from pattern‑matching to uncovering causal mechanisms in life sciences. Researchers showcased interpretable deep‑learning models for RNA splicing, probed AlphaFold2’s confidence, and revealed hidden biases in genomic datasets. New computational engines such as Foldseek‑multimer and protein language models are mapping the uncharted protein universe, while spatial AI tools reconstruct three‑dimensional tissue architecture from sparse data. Clinical applications were front‑and‑center, with AI systems already diagnosing biomarkers from histology and integrating multimodal patient data to forecast outcomes.

EMBL‑EBI’s MGnify platform has expanded its biome catalogue resource to cover 18 distinct environments, now housing over half a million microbial genomes, including a human gut collection of nearly 300,000 genomes. The catalogues combine metagenome‑assembled genomes and isolate sequences, providing...

Researchers from EMBL Heidelberg and partner institutions have created the first detailed map of immune and stromal cell organization within human lymph nodes, revealing how this architecture deteriorates in lymphomas. They discovered an inflammatory vicious cycle where T‑cell interferon signals...

Stephanie Lo has been appointed Protein Function Content Team Leader at EMBL‑EBI, overseeing the curation of protein function data for UniProt. She brings experience from leading the Global Pneumococcal Sequencing project at the Wellcome Sanger Institute, where she linked bacterial...

A cross‑disciplinary study from EMBL and the University of Geneva demonstrates that variation in tissue mechanical properties—coined “mechanotypes”—drives the striking body‑shape diversity of cnidarians. By integrating active‑surface theory with experiments on six species, the team identified three mechanical modules that...

Senior Ensembl Outreach Officer Aleña Mushtaq leverages her molecular biology background to deliver global training on the Ensembl genome browser. Ensembl, a free 26‑year‑old platform, integrates gene, variation and regulatory data, supporting projects like the Darwin Tree of Life. Mushtaq’s...

The International Nucleotide Sequence Database Collaboration (INSDC) has issued new minimal specifications to modernise how nucleotide sequence data and metadata are submitted and exchanged. The framework outlines supported data types, required metadata, linkage rules, and quality checks, creating a unified...

A joint effort by EMBL‑EBI, Google DeepMind, NVIDIA and Seoul National University has added millions of AI‑predicted protein complex structures to the AlphaFold Database, marking the largest collection of such data to date. The release focuses on high‑confidence homodimers, delivering...
EMBL‑EBI has launched BioAIrepo, a dedicated repository that makes life‑science machine‑learning models FAIR—findable, accessible, interoperable and reusable. The pilot catalogue aggregates imaging and genomics models from the BioImage Model Zoo and Kipoi, providing code, weights, training data and citation metadata....