AlphaFold is a deep learning system developed by DeepMind that predicts the three-dimensional structure of a protein from its amino acid sequence with accuracy approaching that of experimental methods such as X-ray crystallography, a problem known as the protein folding problem. It uses an attention-based neural network architecture trained on known protein structures and evolutionary sequence data to output per-residue coordinates along with a confidence estimate. AlphaFold’s predictions are widely used in computational biology and genomics research, and its structure database has substantially accelerated work in drug discovery and molecular biology.