Introducing AlphaFold 3, a new AI model was developed by Google DeepMind and Isomorphic Labs. We hope it will transform our understanding of the biological world and drug discovery by accurately predicting the structure and interaction of proteins, DNA, RNA, ligands, and more.
Billions of molecular machines are inside every plant, animal, and human cell. They’re made up of proteins, DNA, and other molecules, but no single piece works independently. Only by seeing how they interact together across millions of combinations can we start to truly understand life’s processes.
In a paper published in Nature, we introduce AlphaFold 3, a revolutionary model that can predict the structure and interactions of all life’s molecules with unprecedented accuracy. For the interactions of proteins with other molecule types, we see at least a 50% improvement compared with existing prediction methods. For some important categories of interaction, we have doubled prediction accuracy.
AlphaFold 3 will help transform our understanding of the biological world and drug discovery. Scientists can access most of its capabilities for free through our newly launched AlphaFold Server, an easy-to-use research tool. To build on AlphaFold 3’s potential for drug design, Isomorphic Labs is already collaborating with pharmaceutical companies to apply it to real-world drug design challenges and, ultimately, develop new life-changing treatments for patients.
Our new model builds on the foundations of AlphaFold 2, which in 2020 made a fundamental breakthrough in protein structure prediction. So far, millions of researchers globally have used AlphaFold 2 to make discoveries in malaria vaccines, cancer treatments, and enzyme design. AlphaFold has been cited more than 20,000 times, and its scientific impact is recognized through many prizes, most recently the Breakthrough Prize in Life Sciences. AlphaFold 3 takes us beyond proteins to a broad spectrum of biomolecules. This leap could unlock more transformative science, from developing renewable materials and more resilient crops to accelerating drug design and genomics research.
How AlphaFold 3 reveals life’s molecules
AlphaFold 3 generates its joint 3D structure with an input list of molecules, revealing how they all fit together. It models large biomolecules such as proteins, DNA, and RNA and small molecules, also known as ligands—a category encompassing many drugs. Furthermore, AlphaFold 3 can model chemical modifications to these molecules, which control the healthy functioning of cells that, when disrupted, can lead to disease.
AlphaFold 3’s capabilities come from its next-generation architecture and training that cover all of life’s molecules. At the model’s core is an improved version of our Evoformer module—a deep learning architecture underpinning AlphaFold 2’s incredible performance. After processing the inputs, AlphaFold 3 assembles its predictions using a diffusion network akin to those found in AI image generators. The diffusion process starts with a cloud of atoms and, over many steps, converges on its final, most accurate molecular structure.
AlphaFold 3’s molecular interaction predictions surpass the accuracy of all existing systems. As a single model that computes entire molecular complexes holistically, it can uniquely unify scientific insights.
Leading drug discovery at Isomorphic Labs
AlphaFold 3 creates capabilities for drug design with predictions for molecules commonly used in drugs, such as ligands and antibodies, that bind to proteins to change how they interact in human health and disease.
AlphaFold 3 achieves unprecedented accuracy in predicting drug-like interactions, including binding proteins with ligands and antibodies with their target proteins. It is 50% more accurate than the best traditional methods on the PoseBusters benchmark without needing the input of any structural information, making it the first AI system to surpass physics-based tools for biomolecular structure prediction. Predicting antibody-protein binding is critical to understanding aspects of the human immune response and designing new antibodies, a growing class of therapeutics.
Using AlphaFold 3 combined with a complementary suite of in-house AI models, Isomorphic Labs works on drug design for internal projects and pharmaceutical partners. It is using AlphaFold 3 to accelerate and improve the success of drug design by helping understand how to approach new disease targets and developing novel ways to pursue existing ones that were previously out of reach.
AlphaFold Server: A free and easy-to-use research tool
Google DeepMind’s newly launched AlphaFold Server is the most accurate tool for predicting how proteins interact with other molecules throughout the cell. It is a free platform that scientists worldwide can use for non-commercial research. With just a few clicks, biologists can harness the power of AlphaFold 3 to model structures composed of proteins, DNA, RNA, and a selection of ligands, ions, and chemical modifications.
AlphaFold Server helps scientists make novel hypotheses to test in the lab, speeding up workflows and enabling further innovation. Our platform gives researchers an accessible way to generate predictions, regardless of their access to computational resources or their expertise in machine learning.
Experimental protein-structure prediction can take about the length of a PhD and cost hundreds of thousands of dollars. Our previous model, AlphaFold 2, has been used to predict hundreds of millions of structures, which would have taken hundreds of millions of researcher years at the current rate of experimental structural biology.
With AlphaFold Server, it’s not only about predicting structures anymore, it’s about generously giving access: allowing researchers to ask daring questions and accelerate discoveries.Céline Bouchoux The Francis Crick Institute
Sharing the power of AlphaFold 3 responsibly
With each AlphaFold release, we’ve sought to understand the technology’s broad impact, working with the research and safety community. We take a science-led approach, have conducted extensive assessments to mitigate potential risks and share the widespread benefits to biology and humanity.
Building on the external consultations we carried out for AlphaFold 2, we’ve engaged with more than 50 domain experts and specialist third parties across biosecurity, research, and industry to understand the capabilities of successive AlphaFold models and any potential risks. We also participated in community-wide forums and discussions before AlphaFold 3’s launch.
AlphaFold Server reflects our ongoing commitment to share the benefits of AlphaFold, including our free database of 200 million protein structures. We’ll also be expanding our free AlphaFold education online course with EMBL-EBI and partnerships with organizations in the Global South to equip scientists with the tools they need to accelerate adoption and research, including on underfunded areas such as neglected diseases and food security. We’ll continue working with the scientific community and policymakers to responsibly develop and deploy AI technologies.
Also Read: Explained: TinyLlama – The Promising Generation of Powerful Smaller Language Models
Opening up the future of AI-powered cell biology
AlphaFold 3 brings the biological world into high definition. It allows scientists to see cellular systems in all their complexity across structures, interactions, and modifications. This new window on the molecules of life reveals how they’re all connected and helps understand how those connections affect biological functions — such as the actions of drugs, the production of hormones, and the health-preserving process of DNA repair.
The impacts of AlphaFold 3 and our free AlphaFold Server will be realized through how they empower scientists to accelerate discovery across open questions in biology and new lines of research. We’re just beginning to tap into AlphaFold 3’s potential and can’t wait to see what the future holds.
