← Retour au blog
tech 7 June 2026

Biohub Revolutionizes Protein Biology with a World Model

Biohub has unveiled a world model of protein biology, promising to transform scientific research by mapping the protein universe and designing functional binders with therapeutic-level affinity.

Article inspired by the original source
Biohub releases a world model of protein biology ↗ biohub.org

Biohub and Innovation in Protein Biology

On May 27, 2026, Biohub announced what could redefine the boundaries of protein biology. By unveiling a world model of protein biology, Biohub offers the scientific community a discovery engine that could accelerate research, design, and discovery of new proteins. This model can map proteins across the tree of life, predict their structures, and design new protein binders that work in laboratory experiments.

Why Protein Biology is Crucial

Proteins are essential to almost every function in the human body. They are central to many therapeutic approaches, but designing stable and functional proteins remains a major challenge. By introducing ESMC, ESMFold2, and ESM Atlas, Biohub provides a foundation for modeling the sequence, structure, and function of proteins with unparalleled accuracy.

ESMC: The Cutting-edge Language Model

At the core of this advancement is ESMC, a state-of-the-art language model trained on approximately 2.8 billion sequences from all forms of life. The central scientific hypothesis is that training a language model on these sequences allows it to internalize the fundamental properties that govern protein biology.

ESMFold2: Rapid Structural Transformations

ESMFold2 is designed to transform ESMC's sequence representations into atomically resolved 3D structures of biomolecular complexes. What used to take months now completes in days. The designed binders exhibited high affinity, specificity, and stability, essential properties for clinical use.

ESM Atlas: Massive Navigation in Protein Sequences

ESM Atlas makes ESMC's representations accessible across 6.8 billion protein sequences and 1.1 billion predicted structures. It organizes proteins by relationships the model has learned, revealing previously unseen evolutionary connections. This is the largest application of AI to protein biology to date.

Implications for Research and Medicine

With these advancements, Biohub paves the way for discoveries that could transform medicine. Researchers can now design proteins with clinically useful properties more quickly and efficiently. This could accelerate the development of new treatments for complex diseases like cancer and immune disorders.

Conclusion

Biohub has taken a significant step in protein biology. By making these models available to the scientific community, they empower new heights in biomedical research and development. Let's discuss your project in 15 minutes.

Biohub protein biology AI ESMC ESMFold2
Deepthix newsletter · 100% AI · every Monday 8am

An AI agent reads tech for you.

Our AI agent scans ~200 sources per week and ships the best articles to your inbox Monday 8am. Free. One click to unsubscribe.

Visit the newsletter page →

Want to automate your operations?

Let's talk about your project in 15 minutes.

Book a call