AlphaFold
In plain wordsA DeepMind AI system that predicts the 3D shape of proteins from their chemical sequence, often about as accurately as laboratory experiments. It cracked a problem biologists had worked on for 50 years and won a Nobel Prize.
A series of deep learningdeep learning, n.: Machine learning done with neural networks built from many stacked layers: raw data such as pixels, sound or text goes in at one end, and each layer turns what it receives into something more abstract. systems from DeepMindDeepMind, prop. n.: A London artificial intelligence company, bought by Google in 2014, famous for programs that mastered the game of Go and predicted the shapes of proteins. In 2023 it merged into Google DeepMind. (now Google DeepMindGoogle DeepMind, prop. n.: Google's main artificial intelligence laboratory, formed in 2023 by joining DeepMind and Google Brain. It builds the Gemini models and scientific AI systems such as AlphaFold.) for predicting the three-dimensional structure of proteins. AlphaFold 2, which dominated the CASP14 assessment in November 2020 with a median accuracy score (GDT) of about 92 out of 100, comparable to experimental methods for many targets, takes an amino-acid sequence plus aligned sequences of related proteins, refines representations of residue pairs with an attention-based trunk called the Evoformer, and outputs atomic coordinates with per-residue confidence estimates, trained end to end on structures from the Protein Data Bank. In 2021 DeepMind released the code and, with EMBL's European Bioinformatics Institute, launched a database that by 2022 held predicted structures for over 200 million proteins, nearly every catalogued protein. AlphaFold 3 (May 2024, with Isomorphic Labs) adds a diffusiondiffusion model, n.: An AI that learns to create images, sound or video by practicing noise removal: shown pictures buried under static, it learns to clean them up, then builds new ones starting from pure static.-based module to predict complexes of proteins with DNA, RNA, small molecules and ions.
We used the AlphaFold prediction as a starting model before turning to the cryo-EM data.
Origin
The first AlphaFold won the CASP13 assessment in December 2018. AlphaFold 2 was described by John Jumper and colleagues in Nature in July 2021. Demis HassabisDemis Hassabis, prop. n.: A British AI researcher and former chess prodigy who co-founded DeepMind, led the teams behind AlphaGo and AlphaFold, and shared the 2024 Nobel Prize in Chemistry. and Jumper shared half of the 2024 Nobel Prize in Chemistry for the work.
Usage note
AlphaFold predicts static structures and is less reliable for disordered regions, conformational changes and some complexes; it does not simulate the physical folding process. See Nobel Prizes for AINobel Prizes for AI, phr.: The two 2024 Nobel Prizes, in physics and chemistry, that honored artificial intelligence research for the first time, recognizing neural-network pioneers and the protein-structure system AlphaFold..
Related
- Nobel Prizes for AINobel Prizes for AI, phr.: The two 2024 Nobel Prizes, in physics and chemistry, that honored artificial intelligence research for the first time, recognizing neural-network pioneers and the protein-structure system AlphaFold.
- Google DeepMindGoogle DeepMind, prop. n.: Google's main artificial intelligence laboratory, formed in 2023 by joining DeepMind and Google Brain. It builds the Gemini models and scientific AI systems such as AlphaFold.
- Demis HassabisDemis Hassabis, prop. n.: A British AI researcher and former chess prodigy who co-founded DeepMind, led the teams behind AlphaGo and AlphaFold, and shared the 2024 Nobel Prize in Chemistry.
- deep learningdeep learning, n.: Machine learning done with neural networks built from many stacked layers: raw data such as pixels, sound or text goes in at one end, and each layer turns what it receives into something more abstract.
- diffusion modeldiffusion model, n.: An AI that learns to create images, sound or video by practicing noise removal: shown pictures buried under static, it learns to clean them up, then builds new ones starting from pure static.
- AlphaGoAlphaGo, prop. n.: A DeepMind program that in 2016 beat Lee Sedol, one of the world's best players of the board game Go, a decade or more before many experts expected a computer to manage it.
Sources
- Jumper, J. et al. (2021). Highly Accurate Protein Structure Prediction with AlphaFold. Nature 596, 583–589.
- Abramson, J. et al. (2024). Accurate Structure Prediction of Biomolecular Interactions with AlphaFold 3. Nature 630, 493–500.
Mentioned in
Demis HassabisDemis Hassabis, prop. n.: A British AI researcher and former chess prodigy who co-founded DeepMind, led the teams behind AlphaGo and AlphaFold, and shared the 2024 Nobel Prize in Chemistry., Google DeepMindGoogle DeepMind, prop. n.: Google's main artificial intelligence laboratory, formed in 2023 by joining DeepMind and Google Brain. It builds the Gemini models and scientific AI systems such as AlphaFold., DeepMindDeepMind, prop. n.: A London artificial intelligence company, bought by Google in 2014, famous for programs that mastered the game of Go and predicted the shapes of proteins. In 2023 it merged into Google DeepMind., John JumperJohn Jumper, prop. n.: An American scientist who led Google DeepMind's AlphaFold project, which predicts the 3D shapes of proteins, and shared the 2024 Nobel Prize in Chemistry., Nobel Prizes for AINobel Prizes for AI, phr.: The two 2024 Nobel Prizes, in physics and chemistry, that honored artificial intelligence research for the first time, recognizing neural-network pioneers and the protein-structure system AlphaFold., geometric deep learninggeometric deep learning, n.: An approach to designing neural networks around the shapes and symmetries of the data, whether grids, graphs, surfaces or 3-D structures, so the network respects how that data can be moved or rearranged.