AlphaFold

AlphaFold is DeepMind’s Nobel-recognized system for predicting protein structures, opening a free database of nearly every known protein to researchers worldwide.

Overview

AlphaFold solved a grand challenge of biology: predicting the three-dimensional shape of a protein from its amino-acid sequence. Its successor, AlphaFold 3, extends the same framework to complexes with DNA, RNA, small molecules, and ions, modeling how biomolecules interact rather than just how they fold. The predictions are published through the AlphaFold Protein Structure Database, a free resource that has been accessed by millions of researchers, and the open-source code lets labs run their own predictions. For drug discovery, enzymology, and basic research, structures that once took years of crystallography arrive in minutes, and the system’s accuracy has made structural predictions a routine first step in almost every molecular biology project. Independent assessments such as CASP have repeatedly placed AlphaFold in a class of its own, and the system continues to improve as new training strategies and expanded molecular coverage are released.

Key Features

• Protein structure prediction at experimental accuracy in minutes
• AlphaFold 3 modeling of complexes with DNA, RNA, ligands, and ions
• Free public structure database covering nearly all catalogued proteins
• Open-source code and weights for local and custom predictions
• Colab notebooks that let non-programmers run predictions easily
• Ligand-bound complex predictions useful for early drug-discovery screening

Pricing

AlphaFold is free for academic and commercial use, with the database, code, and model weights openly available. Cloud credits cover larger-scale runs, and DeepMind partners selectively with organizations on specialized applications.

Who Is It For

Structural biologists, drug-discovery teams, and biochemists who need reliable molecular models, and educators who want to show students real structures rather than textbook diagrams. It has effectively become shared infrastructure for the life sciences, much as sequence databases became a generation earlier. Commercial biotech teams also rely on it for early hypothesis testing before committing laboratory resources.