AlphaFold 3: Predicting All Life Molecules' Structures and Interactions

Leading drug discovery at Isomorphic Labs -----------------------------------------

AlphaFold 3 opens up new possibilities for drug design by predicting how molecules commonly used in drugs, like ligands and antibodies, interact with proteins to influence human health and disease.

AlphaFold 3 achieves unmatched accuracy in predicting drug-like interactions, including how proteins bind with ligands and how antibodies interact with their target proteins. It's 50% more accurate than the best traditional methods on the PoseBusters benchmark, and it doesn't need any structural information to start. This makes AlphaFold 3 the first AI system to outdo physics-based tools for biomolecular structure prediction. Understanding antibody-protein binding is crucial for insights into the human immune response and for designing new antibodies, which are increasingly important in medicine.

Isomorphic Labs is using AlphaFold 3, along with other in-house AI models, to enhance drug design for both internal projects and collaborations with pharmaceutical partners. This approach helps us tackle new disease targets and find innovative ways to address existing ones that were previously unattainable.

AlphaFold Server: A free and easy-to-use research tool ------------------------------------------------------ 8AW3 - RNA modifying protein: AlphaFold 3’s prediction for a molecular complex featuring a protein (blue), a strand of RNA (purple), and two ions (yellow) closely matches the true structure (gray). This complex is involved with the creation of other proteins — a cellular process fundamental to life and health.

Google DeepMind's newly launched AlphaFold Server is the world's most accurate tool for predicting how proteins interact with other molecules in the cell. It's a free platform available to scientists worldwide for non-commercial research. With just a few clicks, biologists can use AlphaFold 3 to model structures made up of proteins, DNA, RNA, and a selection of ligands, ions, and chemical modifications.

AlphaFold Server helps scientists generate new hypotheses to test in the lab, speeding up their research and fostering innovation. Our platform makes it easy for researchers to create predictions, no matter their access to computational resources or their expertise in machine learning.

Experimental protein-structure prediction can take as long as a PhD and cost a fortune. Our previous model, AlphaFold 2, has been used to predict hundreds of millions of structures, a task that would have taken hundreds of millions of researcher-years at the current pace of experimental structural biology.