AlphaFold

AlphaFold is an artificial intelligence program developed by DeepMind, a Google-owned company. It is designed to predict the three-dimensional structure of proteins from their amino acid sequence. Predicting protein structures is a long-standing challenge in biology, as the function of a protein is closely related to its shape.

AlphaFold utilizes a deep learning approach, employing neural networks trained on a massive dataset of known protein structures and amino acid sequences. The program's architecture has evolved through different versions, with AlphaFold2 representing a significant leap in accuracy.

The initial AlphaFold system, presented in 2018, achieved competitive results in the 13th Critical Assessment of Structure Prediction (CASP13), a biennial protein structure prediction competition. However, AlphaFold2, presented in 2020 at CASP14, demonstrated significantly improved accuracy, correctly predicting the structure of proteins at a level competitive with experimental methods such as X-ray crystallography and cryo-electron microscopy for many protein targets.

The success of AlphaFold2 has had a major impact on biological research, allowing scientists to quickly and accurately determine the structure of proteins, leading to a better understanding of their function and potential applications in drug discovery and other fields. DeepMind has made the AlphaFold2 code open-source and released a database of predicted protein structures for hundreds of millions of proteins, making it accessible to researchers worldwide. This open access has accelerated research across numerous disciplines, from basic biology to medicine and materials science.