AlphaFold is an artificial intelligence (AI) program developed by DeepMind, a subsidiary of Alphabet, which performs predictions of protein structure. The program is designed as a deep learning system.
AlphaFold AI software has had two major versions. A team of researchers that used AlphaFold 1 (2018) placed first in the overall rankings of the 13th Critical Assessment of Structure Prediction (CASP) in December 2018. The program was particularly successful at predicting the most accurate structure for targets rated as the most difficult by the competition organisers, where no existing template structures were available from proteins with a partially similar sequence.
A team that used AlphaFold 2 (2020) repeated the placement in the CASP competition in November 2020. The team achieved a level of accuracy much higher than any other group. It scored above 90 for around two-thirds of the proteins in CASP's global distance test (GDT), a test that measures the degree to which a computational program predicted structure is similar to the lab experiment determined structure, with 100 being a complete match, within the distance cutoff used for calculating GDT.
AlphaFold 2's results at CASP were described as "astounding" and "transformational." Some researchers noted that the accuracy is not high enough for a third of its predictions, and that it does not reveal the mechanism or rules of protein folding for the protein folding problem to be considered solved. Nevertheless, there has been widespread respect for the technical achievement.
On 15 July 2021 the AlphaFold 2 paper was published at Nature as an advance access publication alongside open source software and a searchable database of species proteomes.
Protein folding problem
Proteins consist of chains of amino acids which spontaneously fold, in a process called protein folding, to form the three dimensional (3-D) structures of the proteins. The 3-D structure is crucial to the biological function of the protein. However, understanding how the amino acid sequence can determine the 3-D structure is highly challenging, and this is called the "protein folding problem". The "protein folding problem" involves understanding the thermodynamics of the interatomic forces that determine the folded stable structure, the mechanism and pathway through which a protein can reach its final folded state with extreme rapidity, and how the native structure of a protein can be predicted from its amino acid sequence.
Protein structures are currently determined experimentally by means of techniques such as X-ray crystallography, cryo-electron microscopy and nuclear magnetic resonance, techniques which are both expensive and time-consuming. Such efforts have identified the structures of about 170,000 proteins over the last 60 years, while there are over 200 million known proteins across all life forms. If it is possible to predict protein structure from the amino-acid sequence alone, it would greatly help to advance scientific research. However, the Levinthal's paradox shows that while a protein can fold in milliseconds, the time it takes to calculate all the possible structures randomly to determine the true native structure is longer than the age of the known universe, which made predicting protein structures a grand challenge in biology for scientists.
Over the years, researchers have applied numerous computational methods to resolve the issue of protein structure prediction, but their accuracy has not been close to experimental techniques except for small simple proteins, thus limiting their value. CASP, which was launched in 1994 to challenge the scientific community to produce their best protein structure predictions, found that GDT scores of only about 40 out of 100 can be achieved for the most difficult proteins by 2016. AlphaFold started competing in the 2018 CASP using an artificial intelligence (AI) deep learning technique.
Algorithm
DeepMind is known to have trained the program on over 170,000 proteins from a public repository of protein sequences and structures. The program uses a form of attention network, a deep learning technique that focuses on having the AI identify parts of a larger problem, then piece it together to obtain the overall solution. The overall training was conducted on processing power between 100 and 200 GPUs. Training the system on this hardware took "a few weeks", after which the program would take "a matter of days" to converge for each structure.