# AlphaFold

> Mediated Wiki article. Canonical URL: https://mediated.wiki/source/AlphaFold
> Markdown URL: https://mediated.wiki/source/AlphaFold.md
> Source: https://en.wikipedia.org/wiki/AlphaFold
> Source revision: 1356897097
> License: Creative Commons Attribution-ShareAlike 4.0 International (https://creativecommons.org/licenses/by-sa/4.0/)

Artificial intelligence program by DeepMind

AlphaFold's predicted structure for [RNA polymerase](/source/RNA_polymerase) T1044 at each layer of the network[1]

Part of a series on Artificial intelligence (AI) Major goals Artificial general intelligence Intelligent agent Recursive self-improvement Planning Computer vision General game playing Knowledge representation Natural language processing Robotics AI safety Approaches Machine learning Symbolic Deep learning Bayesian networks Evolutionary algorithms Neuro-symbolic AI Systems integration Open-source AI data centers Applications Bioinformatics Deepfake Earth sciences Finance Generative AI Art Audio Music Government Healthcare Industry Software development Translation Military Physics Projects Philosophy AI alignment Artificial consciousness The bitter lesson Chinese room Friendly AI Ethics Existential risk Turing test Uncanny valley Human–AI interaction History Timeline Progress AI winter AI boom AI bubble Controversies Deepfake pornography Taylor Swift deepfake pornography controversy Grok sexual deepfake scandal Google Gemini image generation controversy It's the Most Terrible Time of the Year Pause Giant AI Experiments Removal of Sam Altman from OpenAI Statement on AI Risk Tay (chatbot) Théâtre D'opéra Spatial Voiceverse NFT plagiarism scandal Glossary Glossary v t e

**AlphaFold** is an [artificial intelligence](/source/Artificial_intelligence) (AI) program developed by [DeepMind](/source/DeepMind), a subsidiary of [Alphabet](/source/Alphabet_Inc.), which performs [predictions of protein structure](/source/Protein_structure_prediction).[2] It is designed using [deep learning](/source/Deep_learning) techniques.[3]

AlphaFold 1 (2018) placed first in the overall rankings of the 13th [Critical Assessment of Structure Prediction](/source/Critical_Assessment_of_Structure_Prediction) (CASP) in December 2018. It was particularly successful at predicting the most accurate structures for targets rated as most difficult by the competition organizers, where no existing [template structures](/source/Threading_(protein_sequence)) were available from [proteins](/source/Proteins) with partially similar sequences.

AlphaFold 2 (2020) repeated this placement in the CASP14 competition in November 2020.[4] It achieved a level of accuracy much higher than any other entry.[3][5] It scored above 90 on CASP's [global distance test](/source/Global_distance_test) (GDT) for approximately two-thirds of the proteins, a test measuring the similarity between a computationally predicted structure and the experimentally determined structure, where 100 represents a complete match.[3][6] The inclusion of [metagenomic](/source/Metagenomics) data has improved the quality of the prediction of [multiple sequence alignments](/source/Multiple_sequence_alignment). One of the biggest sources of the training data was the custom-built Big Fantastic Database of 65,983,866 protein families, represented as multiple sequence alignments and [Hidden Markov models](/source/Hidden_Markov_model), covering 2,204,359,010 protein sequences from reference databases, metagenomes, and metatranscriptomes.[7]

AlphaFold 2's results at CASP14 were described as "astounding"[8] and "transformational".[9] However, some researchers noted that the accuracy was insufficient for a third of its predictions, and that it did not reveal the underlying mechanism or rules of [protein folding](/source/Protein_folding) for the [protein folding problem](/source/Protein_folding_problem), which remains unsolved.[10][11]

Despite this, the technical achievement was widely recognized. On 15 July 2021, the AlphaFold 2 paper was published in *[Nature](/source/Nature_(journal))* as an advance access publication alongside [open source software](/source/Open-source_software) and a searchable database of species [proteomes](/source/Proteome).[7][12][13] As of November 2025, the paper had been cited nearly 43,000 times.[14]

AlphaFold 3 was announced on 8 May 2024. It can predict the structure of [complexes](/source/Protein_complexes) created by proteins with [DNA](/source/DNA), [RNA](/source/RNA), various [ligands](/source/Ligand_(biochemistry)), and [ions](/source/Ion).[15][16] The new prediction method shows a minimum 50% improvement in accuracy for protein interactions with other molecules compared to existing methods.[17]

[Demis Hassabis](/source/Demis_Hassabis) and [John Jumper](/source/John_M._Jumper) shared one half of the 2024 [Nobel Prize in Chemistry](/source/Nobel_Prize_in_Chemistry), awarded "for protein structure prediction," while the other half went to [David Baker](/source/David_Baker_(biochemist)) "for computational protein design."[18] Hassabis and Jumper had previously won the [Breakthrough Prize in Life Sciences](/source/Breakthrough_Prize_in_Life_Sciences) and the [Albert Lasker Award for Basic Medical Research](/source/Albert_Lasker_Award_for_Basic_Medical_Research) in 2023 for their leadership of the AlphaFold project.[19][20]

## Background

See also: [Protein structure prediction](/source/Protein_structure_prediction) and [De novo protein structure prediction](/source/De_novo_protein_structure_prediction)

Amino-acid chains, known as [polypeptides](/source/Polypeptide), fold to form a protein

[Proteins](/source/Protein) consist of [chains of amino acids](/source/Protein_primary_structure) which [spontaneously fold](/source/Protein_folding) to form the [three dimensional (3-D) structures](/source/Protein_tertiary_structure) of the proteins. The 3-D structure is crucial to understanding the biological function of the protein.

Protein structures can be determined experimentally through techniques such as [X-ray crystallography](/source/X-ray_crystallography), [cryo-electron microscopy](/source/Cryo-Electron_Microscopy) and [nuclear magnetic resonance](/source/Nuclear_magnetic_resonance) (NMR), which are all expensive and time-consuming.[21] Such efforts, using the experimental methods, 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.[6]

Over the years, researchers have applied numerous computational methods to [predict the 3D structures of proteins](/source/Protein_structure_prediction#Ab_initio_protein_modelling) from their amino acid sequences, accuracy of such methods in best possible scenario is close to experimental techniques (NMR) by the use of [homology modeling](/source/Homology_modeling) based on molecular evolution. [CASP](/source/CASP), which was launched in 1994 to challenge the scientific community to produce their best protein structure predictions, found that [GDT](/source/Global_distance_test) scores of only about 40 out of 100 can be achieved for the most difficult proteins by 2016.[6] AlphaFold started competing in the 2018 CASP using an [artificial intelligence](/source/Artificial_intelligence) (AI) [deep learning](/source/Deep_learning) technique.[21]

## Algorithm

DeepMind is known to have trained the program on over 170,000 protein structures from the [Protein Data Bank](/source/Protein_Data_Bank), a public repository of protein sequences and structures. The program uses a form of [attention network](/source/Attention_(machine_learning)), a [deep learning](/source/Deep_learning) technique that focuses on having the [AI](/source/AI) identify parts of a larger problem, then piece it together to obtain the overall solution.[3] The overall training was conducted on processing power between 100 and 200 [GPUs](/source/GPUs).[3]

### AlphaFold 1 (2018)

**AlphaFold 1** (2018) was built on work developed by various teams in the 2010s, work that looked at the large databases of related protein sequences now available from many different organisms (most without known 3D structures), to try to find changes at different [residues](/source/Residue_(chemistry)#Biochemistry) (peptides) that appeared to be correlated, even though the residues were not consecutive in the main chain. Such correlations suggest that the residues may be close to each other physically, even though not close in the sequence, allowing a [contact map](/source/Contact_map) to be estimated. Building on recent work prior to 2018, AlphaFold 1 extended this by estimating a probability distribution for the distances between residues, effectively transforming the contact map into a distance map. It also used more advanced learning methods than previously to develop the inference.[22][23] The code was not made publicly available, except to run on sequences of proteins in the 2018 CASP competition.

### AlphaFold 2 (2020)

AlphaFold 2 performance, experiments, and architecture[24]

Architectural details of AlphaFold 2[24]

The 2020 version of the program (**AlphaFold 2**, 2020) is significantly different from the original version that won CASP 13 in 2018, according to the team at DeepMind.[25][26]

AlphaFold 1 used a number of separately trained modules to produce a guide potential, which was then combined with a physics-based energy potential. AlphaFold 2 replaced this with a system of interconnected sub-networks, forming a single, differentiable, end-to-end model based on pattern recognition. This model was trained in an integrated manner.[26][27] After the neural network's prediction converges, a final refinement step applies local physical constraints using energy minimization based on the [AMBER](/source/AMBER) force field. This step only slightly adjusts the predicted structure.[28]

A key part of the 2020 system are two modules, believed to be based on a [transformer](/source/Transformer_(machine_learning_model)) design, which are used to progressively refine a [vector of information](/source/Word_embedding) for each relationship (or "[edge](/source/Connectivity_(graph_theory))" in graph-theory terminology) between an [amino acid residue](/source/Amino_acid_residue) of the protein and another amino acid residue (these relationships are represented by the array shown in green); and between each amino acid position and each different sequences in the input [sequence alignment](/source/Sequence_alignment) (these relationships are represented by the array shown in red).[27] Internally these refinement transformations contain layers that have the effect of bringing relevant data together and filtering out irrelevant data (the "attention mechanism") for these relationships, in a context-dependent way, learned from training data. These transformations are iterated, the updated information output by one step becoming the input of the next, with the sharpened residue/residue information feeding into the update of the residue/sequence information, and then the improved residue/sequence information feeding into the update of the residue/residue information.[27] As the iteration progresses, according to one report, the "attention algorithm ... mimics the way a person might assemble a jigsaw puzzle: first connecting pieces in small clumps—in this case clusters of amino acids—and then searching for ways to join the clumps in a larger whole."[6][*[needs update](https://en.wikipedia.org/wiki/Wikipedia:Manual_of_Style/Dates_and_numbers#Chronological_items)*]

The output of these iterations then informs the final structure prediction module,[27] which also uses transformers,[29] and is itself then iterated. In an example presented by DeepMind, the structure prediction module achieved a correct topology for the target protein on its first iteration, scored as having a GDT_TS of 78, but with a large number (90%) of stereochemical violations – i.e. unphysical bond angles or lengths. With subsequent iterations the number of stereochemical violations fell. By the third iteration the GDT_TS of the prediction was approaching 90, and by the eighth iteration the number of stereochemical violations was approaching zero.[30]

The training data was originally restricted to single peptide chains. However, the October 2021 update, named AlphaFold-Multimer, included protein complexes in its training data. DeepMind stated this update succeeded about 70% of the time at accurately predicting protein-protein interactions.[31]

### AlphaFold 3 (2024)

Announced on 8 May 2024, **AlphaFold 3** was co-developed by Google DeepMind and [Isomorphic Labs](/source/Isomorphic_Labs), both subsidiaries of [Alphabet](/source/Alphabet_Inc.). AlphaFold 3 is not limited to proteins, as it can also predict the structures of [protein complexes](/source/Protein_complex) with [DNA](/source/DNA), [RNA](/source/RNA), [post-translational modifications](/source/Post-translational_modification) and selected [ligands](/source/Ligand) and [ions](/source/Ion).[32][15]

AlphaFold 3 introduces the "Pairformer," a deep learning architecture inspired by the transformer, which is considered similar to, but simpler than, the Evoformer used in AlphaFold 2.[16] The Pairformer module's initial predictions are refined by a [diffusion model](/source/Diffusion_model). This model begins with a cloud of atoms and iteratively refines their positions, guided by the Pairformer's output, to generate a 3D representation of the molecular structure.[15]

The AlphaFold server was created to provide free access to AlphaFold 3 for non-commercial research.[33] As of November 2025, the AlphaFold 3 research paper has been directly cited more than 9,000 times.[34]

## Competitions

Results achieved for protein prediction by the best reconstructions in the CASP 2018 competition (small circles) and CASP 2020 competition (large circles), compared with results achieved in previous years.
The crimson trend-line shows how a handful of models including AlphaFold 1 achieved a significant step-change in 2018 over the rate of progress that had previously been achieved, particularly in respect of the protein sequences considered the most difficult to predict.
 (Qualitative improvement had been made in earlier years, but it is only as changes bring structures within 8 [Å](/source/Angstrom) of their experimental positions that they start to affect the CASP GDS-TS measure).
 The orange trend-line shows that by 2020 online prediction servers had been able to learn from and match this performance, while the best other groups (green curve) had on average been able to make some improvements on it. However, the black trend curve shows the degree to which AlphaFold 2 had surpassed this again in 2020, across the board.
 The detailed spread of data points indicates the degree of consistency or variation achieved by AlphaFold. Outliers represent the handful of sequences for which it did not make such a successful prediction.

### CASP13

In December 2018, DeepMind's AlphaFold placed first in the overall rankings of the 13th [Critical Assessment of Techniques for Protein Structure Prediction](/source/Critical_Assessment_of_Techniques_for_Protein_Structure_Prediction) (CASP).[35][36]

The program was particularly successfully predicting the most accurate structure for targets rated as the most difficult by the competition organisers, where no existing [template structures](/source/Threading_(protein_sequence)) were available from proteins with a partially similar sequence. AlphaFold gave the best prediction for 25 out of 43 protein targets in this class,[36][37][38] achieving a median score of 58.9 on the CASP's [global distance test](/source/Global_distance_test) (GDT) score, ahead of 52.5 and 52.4 by the two next best-placed teams,[39] who were also using deep learning to estimate contact distances.[40][41] Overall, across all targets, AlphaFold 1 achieved a GDT score of 68.5.[42]

In January 2020, implementations and illustrative code of AlphaFold 1 was released [open-source](/source/Open-source_software) on [GitHub](/source/GitHub).[43][21] but, as stated in the "Read Me" file on that website: "This code can't be used to predict structure of an arbitrary protein sequence. It can be used to predict structure only on the CASP13 dataset (links below). The feature generation code is tightly coupled to our internal infrastructure as well as external tools, hence we are unable to open-source it." Therefore, in essence, the code deposited is not suitable for general use but only for the CASP13 proteins. The company has not announced plans to make their code publicly available as of 5 March 2021.

### CASP14

In November 2020, DeepMind's new version, AlphaFold 2, won CASP14.[44][45] Overall, AlphaFold 2 made the best prediction for 88 out of the 97 targets.[8]

On the competition's preferred [global distance test](/source/Global_distance_test) (GDT) measure of accuracy, the program achieved a median score of 92.4 (out of 100), meaning that more than half of its predictions were scored at better than 92.4% for having their atoms in more-or-less the right place,[46][47] a level of accuracy reported to be comparable to experimental techniques like [X-ray crystallography](/source/X-ray_crystallography).[25][9][42] In 2018 AlphaFold 1 had only reached this level of accuracy in two of all of its predictions.[8] 88% of predictions in the 2020 competition had a GDT_TS score of more than 80. On the group of targets classed as the most difficult, AlphaFold 2 achieved a median score of 87.[*[citation needed](https://en.wikipedia.org/wiki/Wikipedia:Citation_needed)*]

Measured by the [root-mean-square deviation](/source/Root-mean-square_deviation_of_atomic_positions) (RMS-D) of the placement of the alpha-carbon atoms of the protein backbone chain, which tends to be dominated by the performance of the worst-fitted outliers, 88% of AlphaFold 2's predictions had an RMS deviation of less than 4 [Å](/source/Angstrom) for the set of overlapped C-alpha atoms.[8] 76% of predictions achieved better than 3 Å, and 46% had a C-alpha atom RMS accuracy better than 2 Å,[8] with a median RMS deviation in its predictions of 2.1 Å for a set of overlapped CA atoms.[8] AlphaFold 2 also achieved an accuracy in modelling surface [side chains](/source/Side_chain) described as "really really extraordinary".

To further validate AlphaFold 2, the conference organizers approached four leading experimental groups working on structures they found particularly challenging and had been unable to determine. In all four cases the three-dimensional models produced by AlphaFold 2 were sufficiently accurate to determine structures of these proteins by [molecular replacement](/source/Molecular_replacement). These included target T1100 (Af1503), a small [membrane protein](/source/Membrane_protein) studied by experimentalists for ten years.[6]

Of the three structures that AlphaFold 2 had the least success in predicting, one was an unusual [multidomain complex](/source/Protein_domain#Multidomain_proteins) consisting of 52 identical copies of the same [domain](/source/Protein_domain). For all targets with a single domain, excluding only one very large protein and the two structures determined by NMR, AlphaFold 2 achieved a GDT_TS score of over 80.

### CASP15

In 2022, DeepMind did not enter CASP15, but most of the entrants used AlphaFold or tools incorporating AlphaFold.[48]

## Reception

AlphaFold 2 scoring more than 90 in [CASP](/source/CASP)'s [global distance test](/source/Global_distance_test) (GDT) is a great achievement in [computational biology](/source/Computational_biology).[6][9] [Nobel Prize](/source/Nobel_Prize_in_Chemistry) winner and [structural biologist](/source/Structural_biology) [Venki Ramakrishnan](/source/Venki_Ramakrishnan) called the result "a stunning advance on the protein folding problem",[6] adding that "It has occurred decades before many people in the field would have predicted. It will be exciting to see the many ways in which it will fundamentally change biological research."[44]

AlphaFold 2's success received wide media attention.[49][44][50] News pieces appeared in the science press, such as *[Nature](/source/Nature_(journal))*,[9] *[Science](/source/Science_(journal))*,[6] *[MIT Technology Review](/source/MIT_Technology_Review)*,[3] and *[New Scientist](/source/New_Scientist)*,[51][52] and the story was covered by national newspapers.[53][54][55][56] A frequent theme was the ability to predict protein structures based on the constituent amino acid sequence, expected to have benefits in the life sciences—accelerating drug discovery and enabling better understanding of diseases.[9][57] Some have noted that even a perfect answer to the protein *[prediction](/source/Protein_structure_prediction)* problem still leaves questions about the protein *[folding](/source/Protein_folding)* problem (and thus [protein dynamics](/source/Protein_dynamics))—understanding in detail how the folding process actually occurs in nature (and how sometimes they can also [misfold](/source/Proteopathy)).[58]

In 2023, [Demis Hassabis](/source/Demis_Hassabis) and [John Jumper](/source/John_M._Jumper) won the [Breakthrough Prize in Life Sciences](/source/Breakthrough_Prize_in_Life_Sciences)[20] as well as the [Albert Lasker Award for Basic Medical Research](/source/Albert_Lasker_Award_for_Basic_Medical_Research) for their management of the AlphaFold project.[59] Hassabis and Jumper proceeded to win the [Nobel Prize in Chemistry](/source/Nobel_Prize_in_Chemistry) in 2024 for their work on "protein structure prediction" with [David Baker](/source/David_Baker_(biochemist)) of the University of Washington.[19][60]

## Source code

Open access to source code of several AlphaFold versions (excluding AlphaFold 3) has been provided by DeepMind in 2022 after requests from the scientific community.[61][62][63] The source code and weights of AlphaFold 3 were [made available for non-commercial use](/source/Source-available_software) to the scientific community upon request in November 2024. It became publicly available in February 2025, still retaining the non-commercial restriction.[64]

### Clones and derivatives

A number of AlphaFold clones have also been published, mostly with [permissive license](/source/Permissive_software_license) terms. Clones for AlphaFold3 include [ByteDance](/source/ByteDance)'s Protenix ([Apache 2.0 License](/source/Apache_2.0_License)),[65] AlQuraishi Laboratory's OpenFold-3 ([MIT license](/source/MIT_license)), and Boltz-1/2 (MIT license).[66]

There are also clones for older versions, though they became less relevant with the open-source release of AlphaFold 1 and 2 source codes. Still relevant are models, both open- and closed-source, that include modifications to the AlphaFold architecture. For AlphaFold 2, a notable example is ESMFold from [Meta](/source/Meta_Platforms), which replaces the multiple sequence alignment with the latent space of a protein language model.[67]

Open-source tools that complement AlphaFold have also been made. One well-cited example is ColabFold, which uses MMseqs2[68] instead of HHblits to speed up the sequence search, allowing the AlphaFold pipelines to run quickly on [Google Colab](/source/Google_Colab).[69]

## Database of protein models generated by AlphaFold

AlphaFold Protein Structure Database Content Data types captured protein structure prediction Organisms all UniProt proteomes Contact Research center EMBL-EBI Primary citation [7] Access Website https://www.alphafold.ebi.ac.uk/ Download URL yes Tools Web yes Miscellaneous License CC-BY 4.0 Curation policy automatic

The **AlphaFold Protein Structure Database** (AFDB), a joint project between AlphaFold and [EMBL-EBI](/source/EMBL-EBI), was launched on July 22, 2021. At launch, the database contained AlphaFold 1-predicted [models](/source/Protein_structure_prediction) for nearly the complete [UniProt](/source/UniProt) [proteome](/source/Proteome) of humans and 20 [model organisms](/source/Model_organisms), totaling over 365,000 proteins. The database does not include proteins with fewer than 16 or more than 2700 [amino acid residues](/source/Amino_acid_residues),[70] but for humans they are available in the whole batch file.[71] AlphaFold's initial goal (as of early 2022) was to expand the database to cover most of the UniRef90 set, which contains over 100 million proteins. As of May 15, 2022, the database contained 992,316 predictions.[72]

In July 2021, UniProt-KB and [InterPro](/source/InterPro)[73] has been updated to show AlphaFold predictions when available.[74]

On July 28, 2022, the team uploaded to the database the structures of around 200 million proteins from 1 million species, covering nearly every known protein on the planet.[75] The number as of 2024 is 214 million, with 26 million being duplicates (exact sequence matches) of another protein in the database. The predicted structures can differ significantly between duplicates.[76]

As of 2025, the AFDB uses AlphaFold 2 for its predictions. All structures produced remain monomeric, but multimeric structures produced by other databases are linked on the page through the 3D-Beacons API. Foldseek, which provides fast and accurate structure searches, is also integrated. Information from AlphaMissense (a tool that uses AlphaFold to predict the outcome of missense mutations) is also integrated.[77]

### Derived databases

AlphaFill adds cofactors to AlphaFold models where appropriate. This is achieved by searching the [Protein Data Bank](/source/Protein_Data_Bank) for similar structures and transplanting cofactors to analogous positions.[78] It is also linked to by UniProt.

TmAlphaFold docks AlphaFold models to biological membranes, similar to what OPM does for PDB structures.[79]

AFTM uses AlphaFold models to identify transmembrane regions in human proteins, similar to what PDBTM does for PDB structures.[79]

ChannelsDB 2.0 uses PDB or AlphaFill models to calculate the pathway a molecule may take to reach an enzyme's active site or to reach another side of a transmembrane pore.[80]

The AFDB is not updated with UniProt sequences changes. AlphaSync keeps the AFDB in sync with UniProt entry changes, generating updated structures, residue-level features and contacts. It tries to use an AFDB entry for the exact updated sequence when available and run AlphaFold 2 independently otherwise. It fills in AFDB's blank for large (> 2,700 aa) proteins and proteins with special FASTA characters such as B, Z, U or X.[81]

The Encyclopedia of Domains (TED) applies the domain-recognition method from [CATH database](/source/CATH_database) to 188 million unique structures from the AFDB, identifying nearly 365 million domains, which is 100 million more than what sequence-based methods could identify.[76]

The [Evolutionary Classification of Protein Domains](/source/Evolutionary_Classification_of_Protein_Domains) database (ECOD) assigns ECOD classifications to all SwissProt proteins in AFDB.[82]

### Unrelated AlphaFold-based databases

isoform.io is a database of AlphaFold2-generated structures of proposed [splice isoforms](/source/Splice_isoform) in the human genome. It includes information from 237,275 human transcripts. It has been used to detect errors in the mRNA predictions for a handful of genes.[83]

## Performance, validations and limitations

AlphaFold has shown certain limitations.

### AlphaFold 1, 2, and AlphaFold DB

- AlphaFold DB provides models of individual protein chains (monomers), rather than their biologically relevant [complexes](/source/Protein_complexes).[84]

- Many protein regions are predicted with low confidence score, including the [intrinsically disordered protein](/source/Intrinsically_disordered_protein) regions.[85]

- Alphafold-2 was validated for predicting effects of point mutations on structure[86] and free energy,[87] with a partial success.

### AlphaFold 3

- Across several benchmarks, AlphaFold3 has demonstrated, on average, superior performance to conventional search-based docking algorithms in predicting small-molecule–protein binding modes.[88]

- AlphaFold 3 version can predict structures of protein complexes with a very limited set of selected [cofactors](/source/Cofactor_(biochemistry)) and co- and [post-translational modifications](/source/Post-translational_modification).[89] Between 50% and 70% of the structures of the human proteome are incomplete without covalently-attached glycans.[90][84]

- Studies have shown that although AlphaFold3 can jointly model protein–ligand co-folding, its accuracy drops markedly on test cases with low similarity to its training data—an area of particular importance for drug discovery.[91] Other work has found that AlphaFold is insensitive to adversarial decoys generated by altering the physicochemical properties of binding pockets, suggesting potential reliance on training-set memorization rather than genuine chemical awareness.[92]

### General

- In the algorithm, the residues are moved freely, without any restraints. Therefore, during modeling the integrity of the chain is not maintained. As a result, AlphaFold may produce topologically wrong results, like structures with an arbitrary number of knots. (The study uses AlphaFold 2.3.2.)[93]

- The model relies, to some extent, on co-evolutionary information from similar proteins. Therefore, it may not perform as well on synthetic proteins or proteins with very low homology to those in the training database.[94] Benchmarks support this limitation: when applied to naturally evolved de novo proteins, AlphaFold2 often yields low-confidence and predictor-dependent models, and protein language model–based (alignment-free) structure predictors can perform better for orphan proteins than AlphaFold2.[95][96] More broadly, comparative analyses show that structure/disorder predictors (including AlphaFold2 and ESMFold) behave differently on de novo and random-sequence proteins than on conserved proteins, and that confidence metrics can show different relationships with predicted disorder in these sequence classes.[97][98]

- The model's ability to predict multiple [native](/source/Native_state) conformations of proteins is limited.

- Proteins are inherently dynamic, and accessing multiple native conformations is often crucial for understanding their function. However, the model has limited capability to represent these alternative conformational states, particularly those that coexist or interconvert in biological environments.

## Applications

See also: [Earth BioGenome Project](/source/Earth_BioGenome_Project)

AlphaFold has been used to predict structures of proteins of [SARS-CoV-2](/source/Severe_acute_respiratory_syndrome_coronavirus_2), the causative agent of [COVID-19](/source/COVID-19_pandemic). The structures of these proteins were pending experimental detection in early 2020.[99][9] Results were reviewed by scientists at the [Francis Crick Institute](/source/Francis_Crick_Institute) in the United Kingdom before being released to the broader research community. The team also confirmed accurate prediction against the experimentally determined SARS-CoV-2 [spike protein](/source/Coronavirus_spike_protein) that was shared in the [Protein Data Bank](/source/Protein_Data_Bank), an international open-access database, before releasing the computationally determined structures of the under-studied protein molecules.[100] The team acknowledged that although these protein structures might not be the subject of ongoing therapeutical research efforts, they will add to the community's understanding of the SARS-CoV-2 virus.[100] Specifically, AlphaFold 2's prediction of the structure of the *[ORF3a](/source/ORF3a)* protein was very similar to the structure determined by researchers at [University of California, Berkeley](/source/University_of_California%2C_Berkeley) using [cryo-electron microscopy](/source/Cryo-Electron_Microscopy). This specific protein is believed to assist the virus in breaking out of the host cell once it replicates. This protein is also believed to play a role in triggering the inflammatory response to the infection.[101]

## Published works

- Andrew W. Senior *et al.* (December 2019), ["Protein structure prediction using multiple deep neural networks in the 13th Critical Assessment of Protein Structure Prediction (CASP13)"](https://onlinelibrary.wiley.com/doi/abs/10.1002/prot.25834), *Proteins: Structure, Function, Bioinformatics* **87**(12) 1141–1148 [doi](/source/Doi_(identifier)):[10.1002/prot.25834](https://doi.org/10.1002%2Fprot.25834)

- Andrew W. Senior *et al.* (15 January 2020), ["Improved protein structure prediction using potentials from deep learning"](https://www.nature.com/articles/s41586-019-1923-7), *[Nature](/source/Nature_(magazine))* **577** 706–710 [doi](/source/Doi_(identifier)):[10.1038/s41586-019-1923-7](https://doi.org/10.1038%2Fs41586-019-1923-7)

- John Jumper *et al.* (December 2020), "High Accuracy Protein Structure Prediction Using Deep Learning", in *[Fourteenth Critical Assessment of Techniques for Protein Structure Prediction (Abstract Book)](https://predictioncenter.org/casp14/doc/CASP14_Abstracts.pdf)*, pp. 22–24

- John Jumper *et al.* (December 2020), "[AlphaFold 2](https://predictioncenter.org/casp14/doc/presentations/2020_12_01_TS_predictor_AlphaFold2.pdf)". Presentation given at CASP 14.

- Abramson, J., Adler, J., Dunger, J. et al. (May 2024), "[Accurate structure prediction of biomolecular interactions with AlphaFold 3](https://www.nature.com/articles/s41586-024-07487-w)", Nature 630, 493–500 (2024)

## See also

- [Folding@home](/source/Folding%40home)

- [IBM Blue Gene](/source/IBM_Blue_Gene)

- [Foldit](/source/Foldit)

- [Rosetta@home](/source/Rosetta%40home)

- [Human Proteome Folding Project](/source/Human_Proteome_Folding_Project)

- [Predicted Aligned Error](/source/Predicted_Aligned_Error)

## References

1. **[^](#cite_ref-1)** authors), Kathryn Tunyasuvunakool, Jonas Adler, Zachary Wu, Tim Green, Michal Zielinski, Augustin Žídek, Alex Bridgland, Andrew Cowie, Clemens Meyer, Agata Laydon, Sameer Velanka *, Gerard J. Kleywegt *, Alex Bateman *, Richard Evans, Alexander Pritzel, Michael Figurnov, Olaf Ronneberger, Russ Bates, Simon A. A. Kohl, Anna Potapenko, Andrew J. Ballard, Bernardino Romera-Paredes, Stanislav Nikolov, Rishub Jain, Ellen Clancy, David Reiman, Stig Petersen, Andrew Senior, Koray Kavukcuoglu, Ewan Birney *, Pushmeet Kohli, John Jumper, Demis Hassabis (* External (2021-07-22). ["Enabling high-accuracy protein structure prediction at the proteome scale"](https://deepmind.google/blog/enabling-high-accuracy-protein-structure-prediction-at-the-proteome-scale/). *Google DeepMind*. Retrieved 2026-04-11.{{[cite web](https://en.wikipedia.org/wiki/Template:Cite_web)}}: CS1 maint: multiple names: authors list ([link](https://en.wikipedia.org/wiki/Category:CS1_maint:_multiple_names:_authors_list))

1. **[^](#cite_ref-2)** ["AlphaFold"](https://deepmind.com/research/case-studies/alphafold). *Deepmind*. [Archived](https://web.archive.org/web/20210119220351/https://deepmind.com/research/case-studies/alphafold) from the original on 19 January 2021. Retrieved 30 November 2020.

1. ^ [***a***](#cite_ref-mittr20201130_3-0) [***b***](#cite_ref-mittr20201130_3-1) [***c***](#cite_ref-mittr20201130_3-2) [***d***](#cite_ref-mittr20201130_3-3) [***e***](#cite_ref-mittr20201130_3-4) [***f***](#cite_ref-mittr20201130_3-5) ["DeepMind's protein-folding AI has solved a 50-year-old grand challenge of biology"](https://www.technologyreview.com/2020/11/30/1012712/deepmind-protein-folding-ai-solved-biology-science-drugs-disease/). *MIT Technology Review*. [Archived](https://web.archive.org/web/20210828104639/https://www.technologyreview.com/2020/11/30/1012712/deepmind-protein-folding-ai-solved-biology-science-drugs-disease/) from the original on 2021-08-28. Retrieved 2020-11-30.

1. **[^](#cite_ref-cnbc20201130_4-0)** Shead, Sam (2020-11-30). ["DeepMind solves 50-year-old 'grand challenge' with protein folding A.I."](https://www.cnbc.com/2020/11/30/deepmind-solves-protein-folding-grand-challenge-with-alphafold-ai.html) *CNBC*. [Archived](https://web.archive.org/web/20210128182507/https://www.cnbc.com/2020/11/30/deepmind-solves-protein-folding-grand-challenge-with-alphafold-ai.html) from the original on 2021-01-28. Retrieved 2020-11-30.

1. **[^](#cite_ref-Stoddart_5-0)** Stoddart, Charlotte (1 March 2022). ["Structural biology: How proteins got their close-up"](https://knowablemagazine.org/article/living-world/2022/structural-biology-how-proteins-got-their-closeup). *Knowable Magazine*. [doi](/source/Doi_(identifier)):[10.1146/knowable-022822-1](https://doi.org/10.1146%2Fknowable-022822-1). [S2CID](/source/S2CID_(identifier)) [247206999](https://api.semanticscholar.org/CorpusID:247206999). [Archived](https://web.archive.org/web/20220407231646/https://knowablemagazine.org/article/living-world/2022/structural-biology-how-proteins-got-their-closeup) from the original on 7 April 2022. Retrieved 25 March 2022.

1. ^ [***a***](#cite_ref-science20201130_6-0) [***b***](#cite_ref-science20201130_6-1) [***c***](#cite_ref-science20201130_6-2) [***d***](#cite_ref-science20201130_6-3) [***e***](#cite_ref-science20201130_6-4) [***f***](#cite_ref-science20201130_6-5) [***g***](#cite_ref-science20201130_6-6) [***h***](#cite_ref-science20201130_6-7) Robert F. Service, ['The game has changed.' AI triumphs at solving protein structures](https://www.science.org/content/article/game-has-changed-ai-triumphs-solving-protein-structures) [Archived](https://web.archive.org/web/20230624175710/https://www.science.org/content/article/game-has-changed-ai-triumphs-solving-protein-structures) 2023-06-24 at the [Wayback Machine](/source/Wayback_Machine), *[Science](/source/Science_(magazine))*, 30 November 2020

1. ^ [***a***](#cite_ref-nature20210715_7-0) [***b***](#cite_ref-nature20210715_7-1) [***c***](#cite_ref-nature20210715_7-2) Jumper, John; Evans, Richard; Pritzel, Alexander; Green, Tim; Figurnov, Michael; Ronneberger, Olaf; Tunyasuvunakool, Kathryn; Bates, Russ; Žídek, Augustin; Potapenko, Anna; Bridgland, Alex; Meyer, Clemens; Kohl, Simon A A; Ballard, Andrew J; Cowie, Andrew; Romera-Paredes, Bernardino; Nikolov, Stanislav; Jain, Rishub; Adler, Jonas; Back, Trevor; Petersen, Stig; Reiman, David; Clancy, Ellen; Zielinski, Michal; Steinegger, Martin; Pacholska, Michalina; Berghammer, Tamas; Bodenstein, Sebastian; Silver, David; Vinyals, Oriol; Senior, Andrew W; Kavukcuoglu, Koray; Kohli, Pushmeet; Hassabis, Demis (2021-07-15). ["Highly accurate protein structure prediction with AlphaFold"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8371605). *Nature*. **596** (7873): 583–589. [Bibcode](/source/Bibcode_(identifier)):[2021Natur.596..583J](https://ui.adsabs.harvard.edu/abs/2021Natur.596..583J). [doi](/source/Doi_(identifier)):[10.1038/s41586-021-03819-2](https://doi.org/10.1038%2Fs41586-021-03819-2). [PMC](/source/PMC_(identifier)) [8371605](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8371605). [PMID](/source/PMID_(identifier)) [34265844](https://pubmed.ncbi.nlm.nih.gov/34265844).

1. ^ [***a***](#cite_ref-AlQuraishiTweet_8-0) [***b***](#cite_ref-AlQuraishiTweet_8-1) [***c***](#cite_ref-AlQuraishiTweet_8-2) [***d***](#cite_ref-AlQuraishiTweet_8-3) [***e***](#cite_ref-AlQuraishiTweet_8-4) [***f***](#cite_ref-AlQuraishiTweet_8-5) Mohammed AlQuraishi, [CASP14 scores just came out and they're astounding](https://twitter.com/MoAlQuraishi/status/1333383634649313280) [Archived](https://web.archive.org/web/20220804024406/https://twitter.com/MoAlQuraishi/status/1333383634649313280) 2022-08-04 at the [Wayback Machine](/source/Wayback_Machine), Twitter, 30 November 2020.

1. ^ [***a***](#cite_ref-:5_9-0) [***b***](#cite_ref-:5_9-1) [***c***](#cite_ref-:5_9-2) [***d***](#cite_ref-:5_9-3) [***e***](#cite_ref-:5_9-4) [***f***](#cite_ref-:5_9-5) Callaway, Ewen (2020-11-30). "'It will change everything': DeepMind's AI makes gigantic leap in solving protein structures". *Nature*. **588** (7837): 203–204. [Bibcode](/source/Bibcode_(identifier)):[2020Natur.588..203C](https://ui.adsabs.harvard.edu/abs/2020Natur.588..203C). [doi](/source/Doi_(identifier)):[10.1038/d41586-020-03348-4](https://doi.org/10.1038%2Fd41586-020-03348-4). [PMID](/source/PMID_(identifier)) [33257889](https://pubmed.ncbi.nlm.nih.gov/33257889). [S2CID](/source/S2CID_(identifier)) [227243204](https://api.semanticscholar.org/CorpusID:227243204).

1. **[^](#cite_ref-curry_10-0)** Stephen Curry, [No, DeepMind has not solved protein folding](http://occamstypewriter.org/scurry/2020/12/02/no-deepmind-has-not-solved-protein-folding/) [Archived](https://web.archive.org/web/20220729105258/http://occamstypewriter.org/scurry/2020/12/02/no-deepmind-has-not-solved-protein-folding/) 2022-07-29 at the [Wayback Machine](/source/Wayback_Machine), Reciprocal Space (blog), 2 December 2020

1. **[^](#cite_ref-11)** Ball, Phillip (9 December 2020). ["Behind the screens of AlphaFold"](https://www.chemistryworld.com/opinion/behind-the-screens-of-alphafold/4012867.article). *Chemistry World*. [Archived](https://web.archive.org/web/20210815072312/https://www.chemistryworld.com/opinion/behind-the-screens-of-alphafold/4012867.article) from the original on 15 August 2021. Retrieved 10 December 2020.

1. **[^](#cite_ref-12)** ["GitHub - deepmind/alphafold: Open source code for AlphaFold"](https://github.com/deepmind/alphafold). *GitHub*. [Archived](https://web.archive.org/web/20210723182218/https://github.com/deepmind/alphafold) from the original on 2021-07-23. Retrieved 2021-07-24.

1. **[^](#cite_ref-13)** ["AlphaFold Protein Structure Database"](https://alphafold.ebi.ac.uk/). *alphafold.ebi.ac.uk*. [Archived](https://web.archive.org/web/20210724013505/https://alphafold.ebi.ac.uk/) from the original on 2021-07-24. Retrieved 2021-07-24.

1. **[^](#cite_ref-14)** ["Google Scholar"](https://scholar.google.com/scholar?cites=6286436358625670901). *scholar.google.com*. Retrieved 2025-05-01.

1. ^ [***a***](#cite_ref-:0_15-0) [***b***](#cite_ref-:0_15-1) [***c***](#cite_ref-:0_15-2) ["AlphaFold 3 predicts the structure and interactions of all of life's molecules"](https://blog.google/technology/ai/google-deepmind-isomorphic-alphafold-3-ai-model/). *Google*. 2024-05-08. [Archived](https://web.archive.org/web/20240509000044/https://blog.google/technology/ai/google-deepmind-isomorphic-alphafold-3-ai-model/) from the original on 2024-05-09. Retrieved 2024-05-09.

1. ^ [***a***](#cite_ref-:8_16-0) [***b***](#cite_ref-:8_16-1) Abramson, Josh; Adler, Jonas; Dunger, Jack; Evans, Richard; Green, Tim; Pritzel, Alexander; Ronneberger, Olaf; Willmore, Lindsay; Ballard, Andrew J.; Bambrick, Joshua; Bodenstein, Sebastian W.; Evans, David A.; Hung, Chia-Chun; O'Neill, Michael; Reiman, David (2024-05-08). ["Accurate structure prediction of biomolecular interactions with AlphaFold 3"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11168924). *Nature*. **630** (8016): 493–500. [Bibcode](/source/Bibcode_(identifier)):[2024Natur.630..493A](https://ui.adsabs.harvard.edu/abs/2024Natur.630..493A). [doi](/source/Doi_(identifier)):[10.1038/s41586-024-07487-w](https://doi.org/10.1038%2Fs41586-024-07487-w). [ISSN](/source/ISSN_(identifier)) [1476-4687](https://search.worldcat.org/issn/1476-4687). [PMC](/source/PMC_(identifier)) [11168924](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11168924). [PMID](/source/PMID_(identifier)) [38718835](https://pubmed.ncbi.nlm.nih.gov/38718835).

1. **[^](#cite_ref-17)** ["Beyond AlphaFold 3: Navigating Future Challenges in Protein Structure Prediction"](https://www.chemh.com/beyond-alphafold-3-navigating-future-challenges-in-protein-structure-prediction/). 2024-05-10. Retrieved 2024-11-29.

1. **[^](#cite_ref-NobelChemistry2024_18-0)** ["Press release: The Nobel Prize in Chemistry 2024"](https://www.nobelprize.org/prizes/chemistry/2024/press-release/). The Royal Swedish Academy of Sciences. 9 October 2024. Retrieved 29 November 2024. The Royal Swedish Academy of Sciences has decided to award the Nobel Prize in Chemistry 2024 with one half to David Baker..."for computational protein design" and the other half jointly to Demis Hassabis... John Jumper..."for protein structure prediction"

1. ^ [***a***](#cite_ref-:4_19-0) [***b***](#cite_ref-:4_19-1) Hunt, Christian Edwards, Katie (9 October 2024). ["Scientists who used AI to 'crack the code' of almost all proteins win Nobel Prize in chemistry"](https://edition.cnn.com/2024/10/09/science/nobel-prize-chemistry-proteins-baker-hassabis-jumper-intl/index.html). *CNN*. [Archived](https://web.archive.org/web/20241010043705/https://edition.cnn.com/2024/10/09/science/nobel-prize-chemistry-proteins-baker-hassabis-jumper-intl/index.html) from the original on 10 October 2024. Retrieved 9 October 2024.{{[cite news](https://en.wikipedia.org/wiki/Template:Cite_news)}}: CS1 maint: multiple names: authors list ([link](https://en.wikipedia.org/wiki/Category:CS1_maint:_multiple_names:_authors_list))

1. ^ [***a***](#cite_ref-:7_20-0) [***b***](#cite_ref-:7_20-1) Knapp, Alex. ["2023 Breakthrough Prizes Announced: Deepmind's Protein Folders Awarded $3 Million"](https://www.forbes.com/sites/alexknapp/2022/09/22/2023-breakthrough-prizes-announced-deepminds-protein-folders-awarded-3-million/). *Forbes*. [Archived](https://web.archive.org/web/20240509171525/https://www.forbes.com/sites/alexknapp/2022/09/22/2023-breakthrough-prizes-announced-deepminds-protein-folders-awarded-3-million/) from the original on 2024-05-09. Retrieved 2024-05-09.

1. ^ [***a***](#cite_ref-:3_21-0) [***b***](#cite_ref-:3_21-1) [***c***](#cite_ref-:3_21-2) ["AlphaFold: Using AI for scientific discovery"](https://deepmind.com/blog/article/AlphaFold-Using-AI-for-scientific-discovery). *Deepmind*. 15 January 2020. [Archived](https://web.archive.org/web/20220307220937/https://deepmind.com/blog/article/AlphaFold-Using-AI-for-scientific-discovery) from the original on 2022-03-07. Retrieved 2020-11-30.

1. **[^](#cite_ref-22)** [Mohammed AlQuraishi](https://en.wikipedia.org/w/index.php?title=Mohammed_AlQuraishi&action=edit&redlink=1) (May 2019), [AlphaFold at CASP13](https://ccsp.hms.harvard.edu/wp-content/uploads/2020/11/AlphaFold-at-CASP13-AlQuraishi.pdf) [Archived](https://web.archive.org/web/20211122013208/https://ccsp.hms.harvard.edu/wp-content/uploads/2020/11/AlphaFold-at-CASP13-AlQuraishi.pdf) 2021-11-22 at the [Wayback Machine](/source/Wayback_Machine), *Bioinformatics*, **35**(22), 4862–4865 [doi](/source/Doi_(identifier)):[10.1093/bioinformatics/btz422](https://doi.org/10.1093%2Fbioinformatics%2Fbtz422). See also Mohammed AlQuraishi (December 9, 2018), [AlphaFold @ CASP13: "What just happened?"](https://moalquraishi.wordpress.com/2018/12/09/alphafold-casp13-what-just-happened/) [Archived](https://web.archive.org/web/20220729130717/https://moalquraishi.wordpress.com/2018/12/09/alphafold-casp13-what-just-happened/) 2022-07-29 at the [Wayback Machine](/source/Wayback_Machine) (blog post). Mohammed AlQuraishi (15 January 2020), [A watershed moment for protein structure prediction](https://www.nature.com/articles/d41586-019-03951-0) [Archived](https://web.archive.org/web/20220623072409/https://www.nature.com/articles/d41586-019-03951-0) 2022-06-23 at the [Wayback Machine](/source/Wayback_Machine), *[Nature](/source/Nature_(journal))* **577**, 627–628 [doi](/source/Doi_(identifier)):[10.1038/d41586-019-03951-0](https://doi.org/10.1038%2Fd41586-019-03951-0)

1. **[^](#cite_ref-23)** [AlphaFold: Machine learning for protein structure prediction](http://fold.it/portal/node/2008706) [Archived](https://web.archive.org/web/20220512085046/http://fold.it/portal/node/2008706) 2022-05-12 at the [Wayback Machine](/source/Wayback_Machine), [Foldit](/source/Foldit), 31 January 2020

1. ^ [***a***](#cite_ref-nature_24-0) [***b***](#cite_ref-nature_24-1) Jumper, John; et al. (August 2021). ["Highly accurate protein structure prediction with AlphaFold"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8371605). *Nature*. **596** (7873): 583–589. [Bibcode](/source/Bibcode_(identifier)):[2021Natur.596..583J](https://ui.adsabs.harvard.edu/abs/2021Natur.596..583J). [doi](/source/Doi_(identifier)):[10.1038/s41586-021-03819-2](https://doi.org/10.1038%2Fs41586-021-03819-2). [ISSN](/source/ISSN_(identifier)) [1476-4687](https://search.worldcat.org/issn/1476-4687). [PMC](/source/PMC_(identifier)) [8371605](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8371605). [PMID](/source/PMID_(identifier)) [34265844](https://pubmed.ncbi.nlm.nih.gov/34265844).

1. ^ [***a***](#cite_ref-economist20201130_25-0) [***b***](#cite_ref-economist20201130_25-1) ["DeepMind is answering one of biology's biggest challenges"](https://www.economist.com/science-and-technology/2020/11/30/deepmind-is-answering-one-of-biologys-biggest-challenges). *The Economist*. 2020-11-30. [ISSN](/source/ISSN_(identifier)) [0013-0613](https://search.worldcat.org/issn/0013-0613). [Archived](https://web.archive.org/web/20201203085450/https://www.economist.com/science-and-technology/2020/11/30/deepmind-is-answering-one-of-biologys-biggest-challenges) from the original on 2020-12-03. Retrieved 2020-11-30.

1. ^ [***a***](#cite_ref-KahnLessons_26-0) [***b***](#cite_ref-KahnLessons_26-1) Jeremy Kahn, [Lessons from DeepMind's breakthrough in protein-folding A.I.](https://fortune.com/2020/12/01/lessons-from-deepminds-a-i-breakthrough-eye-on-ai/) [Archived](https://web.archive.org/web/20220408095912/https://fortune.com/2020/12/01/lessons-from-deepminds-a-i-breakthrough-eye-on-ai/) 2022-04-08 at the [Wayback Machine](/source/Wayback_Machine), *[Fortune](/source/Fortune_(magazine))*, 1 December 2020

1. ^ [***a***](#cite_ref-block_diagram_27-0) [***b***](#cite_ref-block_diagram_27-1) [***c***](#cite_ref-block_diagram_27-2) [***d***](#cite_ref-block_diagram_27-3) See block diagram. Also John Jumper *et al.* (1 December 2020), [AlphaFold 2 presentation](https://predictioncenter.org/casp14/doc/presentations/2020_12_01_TS_predictor_AlphaFold2.pdf) [Archived](https://web.archive.org/web/20220703015739/https://predictioncenter.org/casp14/doc/presentations/2020_12_01_TS_predictor_AlphaFold2.pdf) 2022-07-03 at the [Wayback Machine](/source/Wayback_Machine), slide 10

1. **[^](#cite_ref-Alpha2Abstract_28-0)** John Jumper et al., conference abstract (December 2020)

1. **[^](#cite_ref-29)** The structure module is stated to use a "3-d equivariant transformer architecture" (John Jumper *et al.* (1 December 2020), [AlphaFold 2 presentation](https://predictioncenter.org/casp14/doc/presentations/2020_12_01_TS_predictor_AlphaFold2.pdf) [Archived](https://web.archive.org/web/20220703015739/https://predictioncenter.org/casp14/doc/presentations/2020_12_01_TS_predictor_AlphaFold2.pdf) 2022-07-03 at the [Wayback Machine](/source/Wayback_Machine), slide 12). One design for a transformer network with [SE(3)](/source/Euclidean_group)-[equivariance](/source/Equivariant_map) was proposed in Fabian Fuchs *et al* [SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks](https://arxiv.org/pdf/2006.10503.pdf) [Archived](https://web.archive.org/web/20211007133527/https://arxiv.org/pdf/2006.10503.pdf) 2021-10-07 at the [Wayback Machine](/source/Wayback_Machine), [NeurIPS](/source/NeurIPS) 2020; also [website](https://fabianfuchsml.github.io/se3transformer/) [Archived](https://web.archive.org/web/20220703015807/https://fabianfuchsml.github.io/se3transformer/) 2022-07-03 at the [Wayback Machine](/source/Wayback_Machine). It is not known how similar this may or may not be to what was used in AlphaFold. See also [the blog post](https://moalquraishi.wordpress.com/2020/12/08/alphafold2-casp14-it-feels-like-ones-child-has-left-home/#s3.3) [Archived](https://web.archive.org/web/20201208164545/https://moalquraishi.wordpress.com/2020/12/08/alphafold2-casp14-it-feels-like-ones-child-has-left-home/#s3.3) 2020-12-08 at the [Wayback Machine](/source/Wayback_Machine) by AlQuaraishi on this, or the [more detailed post](https://fabianfuchsml.github.io/alphafold2/) [Archived](https://web.archive.org/web/20220703015630/https://fabianfuchsml.github.io/alphafold2/) 2022-07-03 at the [Wayback Machine](/source/Wayback_Machine) by Fabian Fuchs

1. **[^](#cite_ref-AF2iterations_30-0)** John Jumper *et al.* (1 December 2020), [AlphaFold 2 presentation](https://predictioncenter.org/casp14/doc/presentations/2020_12_01_TS_predictor_AlphaFold2.pdf) [Archived](https://web.archive.org/web/20220703015739/https://predictioncenter.org/casp14/doc/presentations/2020_12_01_TS_predictor_AlphaFold2.pdf) 2022-07-03 at the [Wayback Machine](/source/Wayback_Machine), slides 12 to 20

1. **[^](#cite_ref-31)** Callaway, Ewen (13 April 2022). ["What's next for AlphaFold and the AI protein-folding revolution"](https://doi.org/10.1038%2Fd41586-022-00997-5). *Nature*. **604** (7905): 234–238. [Bibcode](/source/Bibcode_(identifier)):[2022Natur.604..234C](https://ui.adsabs.harvard.edu/abs/2022Natur.604..234C). [doi](/source/Doi_(identifier)):[10.1038/d41586-022-00997-5](https://doi.org/10.1038%2Fd41586-022-00997-5). [PMID](/source/PMID_(identifier)) [35418629](https://pubmed.ncbi.nlm.nih.gov/35418629). [S2CID](/source/S2CID_(identifier)) [248156195](https://api.semanticscholar.org/CorpusID:248156195).

1. **[^](#cite_ref-32)** Metz, Cade (2024-05-08). ["Google Unveils A.I. for Predicting Behavior of Human Molecules"](https://www.nytimes.com/2024/05/08/technology/google-ai-molecules-alphafold3.html). *The New York Times*. [ISSN](/source/ISSN_(identifier)) [0362-4331](https://search.worldcat.org/issn/0362-4331). [Archived](https://web.archive.org/web/20241010043551/https://www.nytimes.com/2024/05/08/technology/google-ai-molecules-alphafold3.html) from the original on 2024-10-10. Retrieved 2024-05-09.

1. **[^](#cite_ref-33)** ["A non-commercial server of AlphaFold-3"](https://golgi.sandbox.google.com/about). [Archived](https://web.archive.org/web/20241010043551/https://golgi.sandbox.google.com/welcome) from the original on 2024-10-10. Retrieved 2024-05-12.

1. **[^](#cite_ref-34)** ["Google Scholar"](https://scholar.google.com/scholar?cites=7030366881110608605). *scholar.google.com*. Retrieved 2025-05-01.

1. **[^](#cite_ref-35)** [Group performance based on combined z-scores](https://predictioncenter.org/casp13/zscores_final.cgi) [Archived](https://web.archive.org/web/20220308055516/https://predictioncenter.org/casp13/zscores_final.cgi) 2022-03-08 at the [Wayback Machine](/source/Wayback_Machine), CASP 13, December 2018. (AlphaFold = Team 043: A7D)

1. ^ [***a***](#cite_ref-Guardian2018_36-0) [***b***](#cite_ref-Guardian2018_36-1) Sample, Ian (2 December 2018). ["Google's DeepMind predicts 3D shapes of proteins"](https://www.theguardian.com/science/2018/dec/02/google-deepminds-ai-program-alphafold-predicts-3d-shapes-of-proteins). *The Guardian*. [Archived](https://web.archive.org/web/20190718162031/https://www.theguardian.com/science/2018/dec/02/google-deepminds-ai-program-alphafold-predicts-3d-shapes-of-proteins) from the original on 18 July 2019. Retrieved 30 November 2020.

1. **[^](#cite_ref-37)** ["AlphaFold: Using AI for scientific discovery"](https://deepmind.com/blog/article/alphafold-casp13). *Deepmind*. January 2020. [Archived](https://web.archive.org/web/20241010043555/https://deepmind.google/discover/blog/alphafold-using-ai-for-scientific-discovery/) from the original on 10 October 2024. Retrieved 30 November 2020.

1. **[^](#cite_ref-38)** Singh, Arunima (2020). ["Deep learning 3D structures"](https://doi.org/10.1038%2Fs41592-020-0779-y). *Nature Methods*. **17** (3): 249. [doi](/source/Doi_(identifier)):[10.1038/s41592-020-0779-y](https://doi.org/10.1038%2Fs41592-020-0779-y). [ISSN](/source/ISSN_(identifier)) [1548-7105](https://search.worldcat.org/issn/1548-7105). [PMID](/source/PMID_(identifier)) [32132733](https://pubmed.ncbi.nlm.nih.gov/32132733). [S2CID](/source/S2CID_(identifier)) [212403708](https://api.semanticscholar.org/CorpusID:212403708).

1. **[^](#cite_ref-39)** See [CASP 13 data tables](https://predictioncenter.org/casp13/results.cgi?view=tb-sel) [Archived](https://web.archive.org/web/20220314052044/https://predictioncenter.org/casp13/results.cgi?view=tb-sel) 2022-03-14 at the [Wayback Machine](/source/Wayback_Machine) for 043 A7D, 322 Zhang, and 089 MULTICOM

1. **[^](#cite_ref-40)** Wei Zheng *et al*,[Deep-learning contact-map guided protein structure prediction in CASP13](https://pubmed.ncbi.nlm.nih.gov/31365149/) [Archived](https://web.archive.org/web/20220122134714/https://pubmed.ncbi.nlm.nih.gov/31365149/) 2022-01-22 at the [Wayback Machine](/source/Wayback_Machine), *Proteins: Structure, Function, and Bioinformatics*, **87**(12) 1149–1164 [doi](/source/Doi_(identifier)):[10.1002/prot.25792](https://doi.org/10.1002%2Fprot.25792); and [slides](https://www.predictioncenter.org/CASP13/doc/presentations/Pred_CASP13_TS_YZhang-Groups_Redacted.pdf) [Archived](https://web.archive.org/web/20220726235133/https://www.predictioncenter.org/CASP13/doc/presentations/Pred_CASP13_TS_YZhang-Groups_Redacted.pdf) 2022-07-26 at the [Wayback Machine](/source/Wayback_Machine)

1. **[^](#cite_ref-41)** Hou, Jie; Wu, Tianqi; Cao, Renzhi; Cheng, Jianlin (2019-04-25). ["Protein tertiary structure modeling driven by deep learning and contact distance prediction in CASP13"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6800999). *Proteins: Structure, Function, and Bioinformatics*. **87** (12). Wiley: 1165–1178. [bioRxiv](/source/BioRxiv_(identifier)) [10.1101/552422](https://doi.org/10.1101%2F552422). [doi](/source/Doi_(identifier)):[10.1002/prot.25697](https://doi.org/10.1002%2Fprot.25697). [ISSN](/source/ISSN_(identifier)) [0887-3585](https://search.worldcat.org/issn/0887-3585). [PMC](/source/PMC_(identifier)) [6800999](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6800999). [PMID](/source/PMID_(identifier)) [30985027](https://pubmed.ncbi.nlm.nih.gov/30985027).

1. ^ [***a***](#cite_ref-:2_42-0) [***b***](#cite_ref-:2_42-1) ["DeepMind Breakthrough Helps to Solve How Diseases Invade Cells"](https://www.bloomberg.com/news/articles/2020-11-30/deepmind-s-alphafold-crosses-threshold-in-solving-protein-riddle). *Bloomberg.com*. 2020-11-30. [Archived](https://web.archive.org/web/20220405104958/https://www.bloomberg.com/news/articles/2020-11-30/deepmind-s-alphafold-crosses-threshold-in-solving-protein-riddle) from the original on 2022-04-05. Retrieved 2020-11-30.

1. **[^](#cite_ref-43)** ["deepmind/deepmind-research"](https://github.com/deepmind/deepmind-research/tree/master/alphafold_casp13). *GitHub*. [Archived](https://web.archive.org/web/20220201014058/https://github.com/deepmind/deepmind-research/tree/master/alphafold_casp13) from the original on 2022-02-01. Retrieved 2020-11-30.

1. ^ [***a***](#cite_ref-DeepMindAlpha2_44-0) [***b***](#cite_ref-DeepMindAlpha2_44-1) [***c***](#cite_ref-DeepMindAlpha2_44-2) ["AlphaFold: a solution to a 50-year-old grand challenge in biology"](https://deepmind.com/blog/article/alphafold-a-solution-to-a-50-year-old-grand-challenge-in-biology). *Deepmind*. 30 November 2020. [Archived](https://web.archive.org/web/20201130153625/https://deepmind.com/blog/article/alphafold-a-solution-to-a-50-year-old-grand-challenge-in-biology) from the original on 30 November 2020. Retrieved 30 November 2020.

1. **[^](#cite_ref-45)** ["DeepMind's protein-folding AI has solved a 50-year-old grand challenge of biology"](https://www.technologyreview.com/2020/11/30/1012712/deepmind-protein-folding-ai-solved-biology-science-drugs-disease/). *MIT Technology Review*. [Archived](https://web.archive.org/web/20210828104639/https://www.technologyreview.com/2020/11/30/1012712/deepmind-protein-folding-ai-solved-biology-science-drugs-disease/) from the original on 28 August 2021. Retrieved 30 November 2020.

1. **[^](#cite_ref-46)** For the GDT_TS measure used, each atom in the prediction scores a quarter of a point if it is within 8 Å (0.80 nm) of the experimental position; half a point if it is within 4 Å, three-quarters of a point if it is within 2 Å, and a whole point if it is within 1 Å.

1. **[^](#cite_ref-47)** To achieve a GDT_TS score of 92.5, mathematically at least 70% of the structure must be accurate to within 1 Å, and at least 85% must be accurate to within 2 Å,

1. **[^](#cite_ref-48)** Callaway, Ewen (2022-12-13). ["After AlphaFold: protein-folding contest seeks next big breakthrough"](https://doi.org/10.1038%2Fd41586-022-04438-1). *Nature*. **613** (7942): 13–14. [doi](/source/Doi_(identifier)):[10.1038/d41586-022-04438-1](https://doi.org/10.1038%2Fd41586-022-04438-1). [PMID](/source/PMID_(identifier)) [36513827](https://pubmed.ncbi.nlm.nih.gov/36513827). [S2CID](/source/S2CID_(identifier)) [254660427](https://api.semanticscholar.org/CorpusID:254660427).

1. **[^](#cite_ref-CASP_release_49-0)** [Artificial intelligence solution to a 50-year-old science challenge could 'revolutionise' medical research](https://predictioncenter.org/casp14/doc/CASP14_press_release.html) [Archived](https://web.archive.org/web/20220424220605/https://predictioncenter.org/casp14/doc/CASP14_press_release.html) 2022-04-24 at the [Wayback Machine](/source/Wayback_Machine) (press release), [CASP](/source/CASP) organising committee, 30 November 2020

1. **[^](#cite_ref-50)** Brigitte Nerlich, [Protein folding and science communication: Between hype and humility](https://blogs.nottingham.ac.uk/makingsciencepublic/2020/12/04/protein-folding-and-science-communication-between-hype-and-humility/) [Archived](https://web.archive.org/web/20220215161037/https://blogs.nottingham.ac.uk/makingsciencepublic/2020/12/04/protein-folding-and-science-communication-between-hype-and-humility/) 2022-02-15 at the [Wayback Machine](/source/Wayback_Machine), [University of Nottingham](/source/University_of_Nottingham) blog, 4 December 2020

1. **[^](#cite_ref-51)** Michael Le Page, [DeepMind's AI biologist can decipher secrets of the machinery of life](https://www.newscientist.com/article/2261156-deepminds-ai-biologist-can-decipher-secrets-of-the-machinery-of-life/) [Archived](https://web.archive.org/web/20220802003315/https://www.newscientist.com/article/2261156-deepminds-ai-biologist-can-decipher-secrets-of-the-machinery-of-life/) 2022-08-02 at the [Wayback Machine](/source/Wayback_Machine), *[New Scientist](/source/New_Scientist)*, 30 November 2020

1. **[^](#cite_ref-52)** [The predictions of DeepMind's latest AI could revolutionise medicine](https://www.newscientist.com/article/2261613-the-predictions-of-deepminds-latest-ai-could-revolutionise-medicine/) [Archived](https://web.archive.org/web/20211107163155/https://www.newscientist.com/article/2261613-the-predictions-of-deepminds-latest-ai-could-revolutionise-medicine/) 2021-11-07 at the [Wayback Machine](/source/Wayback_Machine), *[New Scientist](/source/New_Scientist)*, 2 December 2020

1. **[^](#cite_ref-53)** Tom Whipple, [Deepmind finds biology's 'holy grail' with answer to protein problem](https://www.thetimes.com/edition/news/deepmind-finds-biology-s-holy-grail-with-answer-to-protein-problem-htg6s7qlq), *[The Times](/source/The_Times)* (online), 30 November 2020. Tom Whipple wrote six articles on the subject for *The Times* when the news broke. ([thread](https://twitter.com/whippletom/status/1333494448420958210) [Archived](https://web.archive.org/web/20211108142415/https://twitter.com/whippletom/status/1333494448420958210) 2021-11-08 at the [Wayback Machine](/source/Wayback_Machine)).

1. **[^](#cite_ref-54)** [Cade Metz](https://en.wikipedia.org/w/index.php?title=Cade_Metz&action=edit&redlink=1), [London A.I. Lab Claims Breakthrough That Could Accelerate Drug Discovery](https://www.nytimes.com/2020/11/30/technology/deepmind-ai-protein-folding.html) [Archived](https://web.archive.org/web/20220804151432/https://www.nytimes.com/2020/11/30/technology/deepmind-ai-protein-folding.html) 2022-08-04 at the [Wayback Machine](/source/Wayback_Machine), *[New York Times](/source/New_York_Times)*, 30 November 2020

1. **[^](#cite_ref-55)** Ian Sample,[DeepMind AI cracks 50-year-old problem of protein folding](https://www.theguardian.com/technology/2020/nov/30/deepmind-ai-cracks-50-year-old-problem-of-biology-research) [Archived](https://web.archive.org/web/20201130160106/https://www.theguardian.com/technology/2020/nov/30/deepmind-ai-cracks-50-year-old-problem-of-biology-research) 2020-11-30 at the [Wayback Machine](/source/Wayback_Machine), *[The Guardian](/source/The_Guardian)*, 30 November 2020

1. **[^](#cite_ref-56)** Lizzie Roberts, ['Once in a generation advance' as Google AI researchers crack 50-year-old biological challenge](https://www.telegraph.co.uk/news/2020/11/30/google-ai-researchers-crack-50-year-old-protein-folding-problem/) [Archived](https://web.archive.org/web/20220804003744/https://www.telegraph.co.uk/news/2020/11/30/google-ai-researchers-crack-50-year-old-protein-folding-problem/) 2022-08-04 at the [Wayback Machine](/source/Wayback_Machine). *[Daily Telegraph](/source/Daily_Telegraph)*, 30 November 2020

1. **[^](#cite_ref-57)** [Tim Hubbard](/source/Tim_Hubbard), [The secret of life, part 2: the solution of the protein folding problem.](https://timjph.medium.com/the-secret-of-life-part-2-the-solution-of-the-protein-folding-problem-c544f3a77ee3) [Archived](https://web.archive.org/web/20220514070022/https://timjph.medium.com/the-secret-of-life-part-2-the-solution-of-the-protein-folding-problem-c544f3a77ee3) 2022-05-14 at the [Wayback Machine](/source/Wayback_Machine), [medium.com](/source/Medium.com), 30 November 2020

1. **[^](#cite_ref-58)** e.g. Greg Bowman, [Protein folding and related problems remain unsolved despite AlphaFold's advance](https://foldingathome.org/2020/12/08/protein-folding-and-related-problems-remain-unsolved-despite-alphafolds-advance/) [Archived](https://web.archive.org/web/20220713073427/https://foldingathome.org/2020/12/08/protein-folding-and-related-problems-remain-unsolved-despite-alphafolds-advance/) 2022-07-13 at the [Wayback Machine](/source/Wayback_Machine), [Folding@home](/source/Folding%40home) blog, 8 December 2020

1. **[^](#cite_ref-59)** Sample, Ian (2023-09-21). ["Team behind AI program AlphaFold win Lasker science prize"](https://www.theguardian.com/science/2023/sep/21/team-behind-ai-program-alphafold-win-lasker-science-prize). *The Guardian*. [ISSN](/source/ISSN_(identifier)) [0261-3077](https://search.worldcat.org/issn/0261-3077). [Archived](https://web.archive.org/web/20241010043702/https://www.theguardian.com/science/2023/sep/21/team-behind-ai-program-alphafold-win-lasker-science-prize) from the original on 2024-10-10. Retrieved 2024-05-09.

1. **[^](#cite_ref-60)** ["The Nobel Prize in Chemistry 2024"](https://www.nobelprize.org/prizes/chemistry/2024/press-release/). *NobelPrize.org*. [Archived](https://web.archive.org/web/20241009100757/https://www.nobelprize.org/prizes/chemistry/2024/press-release/) from the original on 2024-10-09. Retrieved 2024-10-10.

1. **[^](#cite_ref-ElPais_61-0)** Domínguez, Nuño (2020-12-02). ["La inteligencia artificial arrasa en uno de los problemas más importantes de la biología"](https://elpais.com/ciencia/2020-12-02/la-inteligencia-artificial-arrasa-en-uno-de-los-problemas-mas-importantes-de-la-biologia.html). *El País* (in Spanish). [Archived](https://web.archive.org/web/20220726121550/https://elpais.com/ciencia/2020-12-02/la-inteligencia-artificial-arrasa-en-uno-de-los-problemas-mas-importantes-de-la-biologia.html) from the original on 2022-07-26. Retrieved 2024-05-12.

1. **[^](#cite_ref-62)** Briggs, David (2020-12-04). ["If Google's Alphafold2 really has solved the protein folding problem, they need to show their working"](https://www.skeptic.org.uk/2020/12/if-googles-alphafold2-really-has-solved-the-protein-folding-problem-they-need-to-show-their-working/). *The Skeptic*. [Archived](https://web.archive.org/web/20240512225437/https://www.skeptic.org.uk/2020/12/if-googles-alphafold2-really-has-solved-the-protein-folding-problem-they-need-to-show-their-working/) from the original on 2024-05-12. Retrieved 2024-05-12.

1. **[^](#cite_ref-HassabisTweet_210618_63-0)** [Demis Hassabis](/source/Demis_Hassabis), ["Brief update on some exciting progress on #AlphaFold!"](https://twitter.com/demishassabis/status/1405922961710854144) [Archived](https://web.archive.org/web/20220722104740/https://twitter.com/demishassabis/status/1405922961710854144) 2022-07-22 at the [Wayback Machine](/source/Wayback_Machine) (tweet), via [twitter](/source/Twitter), 18 June 2021

1. **[^](#cite_ref-64)** [*google-deepmind/alphafold3*](https://github.com/google-deepmind/alphafold3?tab=readme-ov-file#alphafold-3-source-code-and-model-parameters), Google DeepMind, 2025-02-08, retrieved 2025-02-08

1. **[^](#cite_ref-65)** ByteDance AML AI4Science Team; Chen, Xinshi; Zhang, Yuxuan; Lu, Chan; Ma, Wenzhi; Guan, Jiaqi; Gong, Chengyue; Yang, Jincai; Zhang, Hanyu; Zhang, Ke; Wu, Shenghao; Zhou, Kuangqi; Yang, Yanping; Liu, Zhenyu; Wang, Lan; Shi, Bo; Shi, Shaochen; Xiao, Wenzhi (2025-01-11). "Protenix - Advancing Structure Prediction Through a Comprehensive AlphaFold3 Reproduction". [bioRxiv](/source/BioRxiv_(identifier)) [10.1101/2025.01.08.631967](https://doi.org/10.1101%2F2025.01.08.631967).{{[cite bioRxiv](https://en.wikipedia.org/wiki/Template:Cite_bioRxiv)}}: CS1 maint: numeric names: authors list ([link](https://en.wikipedia.org/wiki/Category:CS1_maint:_numeric_names:_authors_list))

1. **[^](#cite_ref-66)** Naddaf, Miryam (28 October 2025). "Open-source protein structure AI aims to match AlphaFold". *Nature*. [doi](/source/Doi_(identifier)):[10.1038/d41586-025-03546-y](https://doi.org/10.1038%2Fd41586-025-03546-y). [PMID](/source/PMID_(identifier)) [41152525](https://pubmed.ncbi.nlm.nih.gov/41152525).

1. **[^](#cite_ref-67)** Lin, Z; Akin, H; Rao, R; Hie, B; Zhu, Z; Lu, W; Smetanin, N; Verkuil, R; Kabeli, O; Shmueli, Y; Dos Santos Costa, A; Fazel-Zarandi, M; Sercu, T; Candido, S; Rives, A (17 March 2023). ["Evolutionary-scale prediction of atomic-level protein structure with a language model"](https://doi.org/10.1126%2Fscience.ade2574). *Science*. **379** (6637): 1123–1130. [Bibcode](/source/Bibcode_(identifier)):[2023Sci...379.1123L](https://ui.adsabs.harvard.edu/abs/2023Sci...379.1123L). [doi](/source/Doi_(identifier)):[10.1126/science.ade2574](https://doi.org/10.1126%2Fscience.ade2574). [PMID](/source/PMID_(identifier)) [36927031](https://pubmed.ncbi.nlm.nih.gov/36927031).

1. **[^](#cite_ref-68)** Steinegger, Martin; Söding, Johannes (2017). ["MMseqs2 enables sensitive protein sequence searching for the analysis of massive data sets"](https://www.nature.com/articles/nbt.3988). *Nature Biotechnology*. **35** (11): 1026–1028. [doi](/source/Doi_(identifier)):[10.1038/nbt.3988](https://doi.org/10.1038%2Fnbt.3988). [ISSN](/source/ISSN_(identifier)) [1087-0156](https://search.worldcat.org/issn/1087-0156). [PMID](/source/PMID_(identifier)) [29035372](https://pubmed.ncbi.nlm.nih.gov/29035372).

1. **[^](#cite_ref-69)** Mirdita, Milot; Schütze, Konstantin; Moriwaki, Yoshitaka; Heo, Lim; Ovchinnikov, Sergey; Steinegger, Martin (2022-05-30). ["ColabFold: Making protein folding accessible to all"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9184281). *Nature Methods*. **19** (6): 679–682. [doi](/source/Doi_(identifier)):[10.1038/s41592-022-01488-1](https://doi.org/10.1038%2Fs41592-022-01488-1). [PMC](/source/PMC_(identifier)) [9184281](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9184281). [PMID](/source/PMID_(identifier)) [35637307](https://pubmed.ncbi.nlm.nih.gov/35637307).

1. **[^](#cite_ref-70)** ["AlphaFold Protein Structure Database"](https://alphafold.ebi.ac.uk/faq). *alphafold.ebi.ac.uk*. [Archived](https://web.archive.org/web/20220729144612/https://alphafold.ebi.ac.uk/Faq) from the original on 2022-07-29. Retrieved 2021-07-29.

1. **[^](#cite_ref-71)** ["AlphaFold Protein Structure Database"](https://alphafold.ebi.ac.uk/download). *alphafold.ebi.ac.uk*. [Archived](https://web.archive.org/web/20220729184651/https://alphafold.ebi.ac.uk/download) from the original on 29 July 2022. Retrieved 27 July 2021.

1. **[^](#cite_ref-72)** ["AlphaFold Protein Structure Database"](https://www.alphafold.ebi.ac.uk/). *www.alphafold.ebi.ac.uk*. [Archived](https://web.archive.org/web/20220802094731/https://www.alphafold.ebi.ac.uk/) from the original on 2022-08-02. Retrieved 2021-07-24.

1. **[^](#cite_ref-73)** InterPro (22 July 2021). ["Alphafold Structure Predictions Available In Interpro"](https://proteinswebteam.github.io/interpro-blog/2021/07/22/AlphaFold-structure-predictions-available-in-InterPro/). *proteinswebteam.github.io*. [Archived](https://web.archive.org/web/20211105155552/https://proteinswebteam.github.io/interpro-blog/2021/07/22/AlphaFold-structure-predictions-available-in-InterPro/) from the original on 2021-11-05. Retrieved 2021-07-29.

1. **[^](#cite_ref-74)** ["Putting the power of AlphaFold into the world's hands"](https://deepmind.com/blog/article/putting-the-power-of-alphafold-into-the-worlds-hands). *Deepmind*. 22 July 2022. [Archived](https://web.archive.org/web/20210724034333/https://deepmind.com/blog/article/putting-the-power-of-alphafold-into-the-worlds-hands) from the original on 24 July 2021. Retrieved 24 July 2021.

1. **[^](#cite_ref-75)** Callaway, Ewen (2022-07-28). ["'The entire protein universe': AI predicts shape of nearly every known protein"](https://doi.org/10.1038%2Fd41586-022-02083-2). *Nature*. **608** (7921): 15–16. [Bibcode](/source/Bibcode_(identifier)):[2022Natur.608...15C](https://ui.adsabs.harvard.edu/abs/2022Natur.608...15C). [doi](/source/Doi_(identifier)):[10.1038/d41586-022-02083-2](https://doi.org/10.1038%2Fd41586-022-02083-2). [PMID](/source/PMID_(identifier)) [35902752](https://pubmed.ncbi.nlm.nih.gov/35902752). [S2CID](/source/S2CID_(identifier)) [251159714](https://api.semanticscholar.org/CorpusID:251159714).

1. ^ [***a***](#cite_ref-Lau24_76-0) [***b***](#cite_ref-Lau24_76-1) Lau, Andy M.; Bordin, Nicola; Kandathil, Shaun M.; Sillitoe, Ian; Waman, Vaishali P.; Wells, Jude; Orengo, Christine A.; Jones, David T. (November 2024). ["Exploring structural diversity across the protein universe with The Encyclopedia of Domains"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7618865). *Science*. **386** (6721) eadq4946. [Bibcode](/source/Bibcode_(identifier)):[2024Sci...386q4946L](https://ui.adsabs.harvard.edu/abs/2024Sci...386q4946L). [doi](/source/Doi_(identifier)):[10.1126/science.adq4946](https://doi.org/10.1126%2Fscience.adq4946). [PMC](/source/PMC_(identifier)) [7618865](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7618865). [PMID](/source/PMID_(identifier)) [39480926](https://pubmed.ncbi.nlm.nih.gov/39480926).

1. **[^](#cite_ref-77)** Fleming, Jennifer; Magana, Paulyna; Nair, Sreenath; Tsenkov, Maxim; Bertoni, Damian; Pidruchna, Ivanna; Lima Afonso, Marcelo Querino; Midlik, Adam; Paramval, Urmila; Žídek, Augustin; Laydon, Agata; Kovalevskiy, Oleg; Pan, Joshua; Cheng, Jun; Avsec, Žiga; Bycroft, Clare; Wong, Lai Hong; Last, Meera; Mirdita, Milot; Steinegger, Martin; Kohli, Pushmeet; Váradi, Mihály; Velankar, Sameer (August 2025). ["AlphaFold Protein Structure Database and 3D-Beacons: New Data and Capabilities"](https://doi.org/10.1016%2Fj.jmb.2025.168967). *Journal of Molecular Biology*. **437** (15) 168967. [doi](/source/Doi_(identifier)):[10.1016/j.jmb.2025.168967](https://doi.org/10.1016%2Fj.jmb.2025.168967). [PMID](/source/PMID_(identifier)) [40133787](https://pubmed.ncbi.nlm.nih.gov/40133787).

1. **[^](#cite_ref-78)** Hekkelman, Maarten L.; de Vries, Ida; Joosten, Robbie P.; Perrakis, Anastassis (February 2023). ["AlphaFill: enriching AlphaFold models with ligands and cofactors"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9911346). *Nature Methods*. **20** (2): 205–213. [Bibcode](/source/Bibcode_(identifier)):[2023NatCB..20..205H](https://ui.adsabs.harvard.edu/abs/2023NatCB..20..205H). [doi](/source/Doi_(identifier)):[10.1038/s41592-022-01685-y](https://doi.org/10.1038%2Fs41592-022-01685-y). [PMC](/source/PMC_(identifier)) [9911346](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9911346). [PMID](/source/PMID_(identifier)) [36424442](https://pubmed.ncbi.nlm.nih.gov/36424442).

1. ^ [***a***](#cite_ref-Rosignoli25_79-0) [***b***](#cite_ref-Rosignoli25_79-1) Rosignoli, Serena; Pacelli, Maddalena; Manganiello, Francesca; Paiardini, Alessandro (February 2025). ["An outlook on structural biology after Alpha Fold: tools, limits and perspectives"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11788754). *FEBS Open Bio*. **15** (2): 202–222. [doi](/source/Doi_(identifier)):[10.1002/2211-5463.13902](https://doi.org/10.1002%2F2211-5463.13902). [PMC](/source/PMC_(identifier)) [11788754](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11788754). [PMID](/source/PMID_(identifier)) [39313455](https://pubmed.ncbi.nlm.nih.gov/39313455).

1. **[^](#cite_ref-80)** Špačková, Anna; Vávra, Ondřej; Raček, Tomáš; Bazgier, Václav; Sehnal, David; Damborský, Jiří; Svobodová, Radka; Bednář, David; Berka, Karel (5 January 2024). ["ChannelsDB 2.0: a comprehensive database of protein tunnels and pores in AlphaFold era"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10767935). *Nucleic Acids Research*. **52** (D1): D413–D418. [doi](/source/Doi_(identifier)):[10.1093/nar/gkad1012](https://doi.org/10.1093%2Fnar%2Fgkad1012). [PMC](/source/PMC_(identifier)) [10767935](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10767935). [PMID](/source/PMID_(identifier)) [37956324](https://pubmed.ncbi.nlm.nih.gov/37956324).

1. **[^](#cite_ref-81)** Lang, Benjamin; Babu, M. Madan (2025). ["AlphaSync is an enhanced AlphaFold structure database synchronized with UniProt"](https://doi.org/10.1038%2Fs41594-025-01719-x). *Nature Structural & Molecular Biology*. **32** (12): 2628–2632. [bioRxiv](/source/BioRxiv_(identifier)) [10.1101/2025.03.12.642845](https://doi.org/10.1101%2F2025.03.12.642845). [doi](/source/Doi_(identifier)):[10.1038/s41594-025-01719-x](https://doi.org/10.1038%2Fs41594-025-01719-x). [PMID](/source/PMID_(identifier)) [41219581](https://pubmed.ncbi.nlm.nih.gov/41219581).

1. **[^](#cite_ref-82)** Schaeffer, RD; Zhang, J; Cong, Q; Grishin, NV (March 2026). ["ECOD: Classification of domains in AFDB Swiss-Prot structure predictions"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048469). *PLOS Computational Biology*. **22** (3) e1013431. [Bibcode](/source/Bibcode_(identifier)):[2026PLSCB..2213431S](https://ui.adsabs.harvard.edu/abs/2026PLSCB..2213431S). [doi](/source/Doi_(identifier)):[10.1371/journal.pcbi.1013431](https://doi.org/10.1371%2Fjournal.pcbi.1013431). [PMC](/source/PMC_(identifier)) [13048469](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048469). [PMID](/source/PMID_(identifier)) [41911251](https://pubmed.ncbi.nlm.nih.gov/41911251).

1. **[^](#cite_ref-83)** Sommer, Markus J; Cha, Sooyoung; Varabyou, Ales; Rincon, Natalia; Park, Sukhwan; Minkin, Ilia; Pertea, Mihaela; Steinegger, Martin; Salzberg, Steven L (15 December 2022). ["Structure-guided isoform identification for the human transcriptome"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9812405). *eLife*. **11** e82556. [doi](/source/Doi_(identifier)):[10.7554/eLife.82556](https://doi.org/10.7554%2FeLife.82556). [PMC](/source/PMC_(identifier)) [9812405](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9812405). [PMID](/source/PMID_(identifier)) [36519529](https://pubmed.ncbi.nlm.nih.gov/36519529).

1. ^ [***a***](#cite_ref-DB-Limitation_84-0) [***b***](#cite_ref-DB-Limitation_84-1) ["What use cases does AlphaFold not support?"](https://www.alphafold.ebi.ac.uk/faq#faq-8). *AlphaFold Protein Structure Database*. [Archived](https://web.archive.org/web/20220729144612/https://alphafold.ebi.ac.uk/Faq#faq-8) from the original on 2022-07-29. Retrieved 2022-05-15.

1. **[^](#cite_ref-85)** [AlphaFold heralds a data-driven revolution in biology and medicine](https://www.nature.com/articles/s41591-021-01533-0) [Archived](https://web.archive.org/web/20241010043626/https://www.nature.com/articles/s41591-021-01533-0) 2024-10-10 at the [Wayback Machine](/source/Wayback_Machine), by [Janet M. Thornton](/source/Janet_M._Thornton), Roman A. Laskowski & Neera Borkakoti, [Nature Medicine](/source/Nature_Medicine), volume 12, pages 1666–1669, 12 October 2021

1. **[^](#cite_ref-:1_86-0)** McBride, John M.; Polev, Konstantin; Abdirasulov, Amirbek; Reinharz, Vladimir; Grzybowski, Bartosz A.; Tlusty, Tsvi (2023-11-20). ["AlphaFold2 Can Predict Single-Mutation Effects"](https://link.aps.org/doi/10.1103/PhysRevLett.131.218401). *Physical Review Letters*. **131** (21) 218401. [arXiv](/source/ArXiv_(identifier)):[2204.06860](https://arxiv.org/abs/2204.06860). [Bibcode](/source/Bibcode_(identifier)):[2023PhRvL.131u8401M](https://ui.adsabs.harvard.edu/abs/2023PhRvL.131u8401M). [doi](/source/Doi_(identifier)):[10.1103/PhysRevLett.131.218401](https://doi.org/10.1103%2FPhysRevLett.131.218401). [ISSN](/source/ISSN_(identifier)) [0031-9007](https://search.worldcat.org/issn/0031-9007). [PMID](/source/PMID_(identifier)) [38072605](https://pubmed.ncbi.nlm.nih.gov/38072605). [Archived](https://web.archive.org/web/20240609035807/https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.131.218401) from the original on 2024-06-09. Retrieved 2023-11-26.

1. **[^](#cite_ref-87)** McBride, John M.; Tlusty, Tsvi (2024-08-26). ["AI-Predicted Protein Deformation Encodes Energy Landscape Perturbation"](https://link.aps.org/doi/10.1103/PhysRevLett.133.098401). *Physical Review Letters*. **133** (9) 098401. [arXiv](/source/ArXiv_(identifier)):[2311.18222](https://arxiv.org/abs/2311.18222). [Bibcode](/source/Bibcode_(identifier)):[2024PhRvL.133i8401M](https://ui.adsabs.harvard.edu/abs/2024PhRvL.133i8401M). [doi](/source/Doi_(identifier)):[10.1103/PhysRevLett.133.098401](https://doi.org/10.1103%2FPhysRevLett.133.098401). [ISSN](/source/ISSN_(identifier)) [0031-9007](https://search.worldcat.org/issn/0031-9007). [PMID](/source/PMID_(identifier)) [39270162](https://pubmed.ncbi.nlm.nih.gov/39270162).

1. **[^](#cite_ref-88)** Abramson, Josh; Adler, Jonas; Dunger, Jack; Evans, Richard; Green, Tim; Pritzel, Alexander; Ronneberger, Olaf; Willmore, Lindsay; Ballard, Andrew J.; Bambrick, Joshua; Bodenstein, Sebastian W.; Evans, David A.; Hung, Chia-Chun; O'Neill, Michael; Reiman, David (June 2024). ["Accurate structure prediction of biomolecular interactions with AlphaFold 3"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11168924). *Nature*. **630** (8016): 493–500. [Bibcode](/source/Bibcode_(identifier)):[2024Natur.630..493A](https://ui.adsabs.harvard.edu/abs/2024Natur.630..493A). [doi](/source/Doi_(identifier)):[10.1038/s41586-024-07487-w](https://doi.org/10.1038%2Fs41586-024-07487-w). [ISSN](/source/ISSN_(identifier)) [1476-4687](https://search.worldcat.org/issn/1476-4687). [PMC](/source/PMC_(identifier)) [11168924](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11168924). [PMID](/source/PMID_(identifier)) [38718835](https://pubmed.ncbi.nlm.nih.gov/38718835).

1. **[^](#cite_ref-auto_89-0)** Bagdonas, Haroldas; Fogarty, Carl A.; Fadda, Elisa; Agirre, Jon (2021-10-29). ["The case for post-predictional modifications in the AlphaFold Protein Structure Database"](https://mural.maynoothuniversity.ie/17434/1/ElisaFaddaAlpha2021.pdf) (PDF). *Nature Structural & Molecular Biology*. **28** (11): 869–870. [doi](/source/Doi_(identifier)):[10.1038/s41594-021-00680-9](https://doi.org/10.1038%2Fs41594-021-00680-9). [ISSN](/source/ISSN_(identifier)) [1545-9985](https://search.worldcat.org/issn/1545-9985). [PMID](/source/PMID_(identifier)) [34716446](https://pubmed.ncbi.nlm.nih.gov/34716446). [S2CID](/source/S2CID_(identifier)) [240228913](https://api.semanticscholar.org/CorpusID:240228913). [Archived](https://web.archive.org/web/20240113191703/https://mural.maynoothuniversity.ie/17434/1/ElisaFaddaAlpha2021.pdf) (PDF) from the original on 2024-01-13. Retrieved 2024-01-13.

1. **[^](#cite_ref-90)** An, Hyun Joo; Froehlich, John W; Lebrilla, Carlito B (2009-10-01). ["Determination of glycosylation sites and site-specific heterogeneity in glycoproteins"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2749913). *Current Opinion in Chemical Biology*. Analytical Techniques/Mechanisms. **13** (4): 421–426. [doi](/source/Doi_(identifier)):[10.1016/j.cbpa.2009.07.022](https://doi.org/10.1016%2Fj.cbpa.2009.07.022). [ISSN](/source/ISSN_(identifier)) [1367-5931](https://search.worldcat.org/issn/1367-5931). [PMC](/source/PMC_(identifier)) [2749913](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2749913). [PMID](/source/PMID_(identifier)) [19700364](https://pubmed.ncbi.nlm.nih.gov/19700364).

1. **[^](#cite_ref-91)** Škrinjar, Peter; Eberhardt, Jérôme; Tauriello, Gerardo; Schwede, Torsten; Durairaj, Janani (2025-08-04). "Have protein-ligand cofolding methods moved beyond memorisation?". [bioRxiv](/source/BioRxiv_(identifier)) [10.1101/2025.02.03.636309](https://doi.org/10.1101%2F2025.02.03.636309).

1. **[^](#cite_ref-92)** Masters, Matthew R.; Mahmoud, Amr H.; Lill, Markus A. (2025-10-06). ["Investigating whether deep learning models for co-folding learn the physics of protein-ligand interactions"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12501370). *Nature Communications*. **16** (1): 8854. [Bibcode](/source/Bibcode_(identifier)):[2025NatCo..16.8854M](https://ui.adsabs.harvard.edu/abs/2025NatCo..16.8854M). [doi](/source/Doi_(identifier)):[10.1038/s41467-025-63947-5](https://doi.org/10.1038%2Fs41467-025-63947-5). [ISSN](/source/ISSN_(identifier)) [2041-1723](https://search.worldcat.org/issn/2041-1723). [PMC](/source/PMC_(identifier)) [12501370](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12501370). [PMID](/source/PMID_(identifier)) [41053181](https://pubmed.ncbi.nlm.nih.gov/41053181).

1. **[^](#cite_ref-93)** Dabrowski-Tumanski, Pawel; Stasiak, Andrzej (7 November 2023). ["AlphaFold Blindness to Topological Barriers Affects Its Ability to Correctly Predict Proteins' Topology"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10672856). *Molecules*. **28** (22): 7462. [doi](/source/Doi_(identifier)):[10.3390/molecules28227462](https://doi.org/10.3390%2Fmolecules28227462). [PMC](/source/PMC_(identifier)) [10672856](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10672856). [PMID](/source/PMID_(identifier)) [38005184](https://pubmed.ncbi.nlm.nih.gov/38005184).

1. **[^](#cite_ref-94)** ["DeepMind's latest AI breakthrough could turbocharge drug discovery"](https://www.fastcompany.com/90584816/deepmind-alphafold-alphabet-ai-proteins-drug-discovery). *Fast Company*. [ISSN](/source/ISSN_(identifier)) [1085-9241](https://search.worldcat.org/issn/1085-9241). [Archived](https://web.archive.org/web/20230124205605/https://www.fastcompany.com/90584816/deepmind-alphafold-alphabet-ai-proteins-drug-discovery) from the original on 2023-01-24. Retrieved 2023-01-24.

1. **[^](#cite_ref-95)** Aubel, Margaux; Eicholt, Lars; Bornberg-Bauer, Erich (2023-03-29). ["Assessing structure and disorder prediction tools for de novo emerged proteins in the age of machine learning"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10126731). *F1000Research*. **12**: 347. [doi](/source/Doi_(identifier)):[10.12688/f1000research.130443.1](https://doi.org/10.12688%2Ff1000research.130443.1). [PMC](/source/PMC_(identifier)) [10126731](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10126731). [PMID](/source/PMID_(identifier)) [37113259](https://pubmed.ncbi.nlm.nih.gov/37113259).

1. **[^](#cite_ref-96)** Michaud, Jennifer M.; Madani, Ali; Fraser, James S. (November 2022). ["A language model beats alphafold2 on orphans"](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9669189). *Nature Biotechnology*. **40** (11): 1576–1577. [doi](/source/Doi_(identifier)):[10.1038/s41587-022-01466-0](https://doi.org/10.1038%2Fs41587-022-01466-0). [PMC](/source/PMC_(identifier)) [9669189](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9669189). [PMID](/source/PMID_(identifier)) [36192635](https://pubmed.ncbi.nlm.nih.gov/36192635).

1. **[^](#cite_ref-97)** Middendorf, Lasse; Eicholt, Lars A. (June 2024). ["Random, de novo, and conserved proteins: How structure and disorder predictors perform differently"](https://doi.org/10.1002%2Fprot.26652). *Proteins*. **92** (6): 757–767. [doi](/source/Doi_(identifier)):[10.1002/prot.26652](https://doi.org/10.1002%2Fprot.26652). [PMID](/source/PMID_(identifier)) [38226524](https://pubmed.ncbi.nlm.nih.gov/38226524).

1. **[^](#cite_ref-98)** Liu, Jing; Yuan, Rongqing; Shao, Wei; Wang, Jitong; Silman, Israel; Sussman, Joel L. (August 2023). ["Do "Newly Born" orphan proteins resemble "Never Born" proteins? A study using three deep learning algorithms"](https://doi.org/10.1002%2Fprot.26496). *Proteins*. **91** (8): 1097–1115. [doi](/source/Doi_(identifier)):[10.1002/prot.26496](https://doi.org/10.1002%2Fprot.26496). [PMID](/source/PMID_(identifier)) [37092778](https://pubmed.ncbi.nlm.nih.gov/37092778).

1. **[^](#cite_ref-99)** ["AI Can Help Scientists Find a Covid-19 Vaccine"](https://www.wired.com/story/opinion-ai-can-help-find-scientists-find-a-covid-19-vaccine/). *Wired*. [ISSN](/source/ISSN_(identifier)) [1059-1028](https://search.worldcat.org/issn/1059-1028). [Archived](https://web.archive.org/web/20220423195503/https://www.wired.com/story/opinion-ai-can-help-find-scientists-find-a-covid-19-vaccine/) from the original on 2022-04-23. Retrieved 2020-12-01.

1. ^ [***a***](#cite_ref-:6_100-0) [***b***](#cite_ref-:6_100-1) ["Computational predictions of protein structures associated with COVID-19"](https://deepmind.com/research/open-source/computational-predictions-of-protein-structures-associated-with-COVID-19). *Deepmind*. 4 August 2020. [Archived](https://web.archive.org/web/20220325185435/https://deepmind.com/research/open-source/computational-predictions-of-protein-structures-associated-with-COVID-19) from the original on 2022-03-25. Retrieved 2020-12-01.

1. **[^](#cite_ref-101)** ["How DeepMind's new protein-folding A.I. is already helping to combat the coronavirus pandemic"](https://fortune.com/2020/11/30/covid-protein-folding-deepmind-ai/). *Fortune*. [Archived](https://web.archive.org/web/20201130204444/https://fortune.com/2020/11/30/covid-protein-folding-deepmind-ai/) from the original on 2020-11-30. Retrieved 2020-12-01.

## Further reading

- Carlos Outeiral, [CASP14: what Google DeepMind's AlphaFold 2 really achieved, and what it means for protein folding, biology and bioinformatics](https://www.blopig.com/blog/2020/12/casp14-what-google-deepminds-alphafold-2-really-achieved-and-what-it-means-for-protein-folding-biology-and-bioinformatics/), Oxford Protein Informatics Group. (3 December)

- Mohammed AlQuraishi, [AlphaFold2 @ CASP14: "It feels like one's child has left home."](https://moalquraishi.wordpress.com/2020/12/08/alphafold2-casp14-it-feels-like-ones-child-has-left-home/) (blog), 8 December 2020

- Mohammed AlQuraishi, [The AlphaFold2 Method Paper: A Fount of Good Ideas](https://moalquraishi.wordpress.com/2021/07/25/the-alphafold2-method-paper-a-fount-of-good-ideas/) (blog), 25 July 2021

## External links

- [AlphaFold-3 web server](https://golgi.sandbox.google.com/about)

- [AlphaFold v2.1 code and links to model](https://github.com/deepmind/alphafold) on [GitHub](/source/GitHub)

- [Open access to protein structure predictions for the human proteome and 20 other key organisms](https://alphafold.ebi.ac.uk/) at [European Bioinformatics Institute](/source/European_Bioinformatics_Institute) (AlphaFold Protein Structure Database)

- [CASP 14](https://predictioncenter.org/casp14/index.cgi) website

- [AlphaFold: The making of a scientific breakthrough](https://www.youtube.com/watch?v=gg7WjuFs8F4), DeepMind, via YouTube.

- [ColabFold](https://github.com/sokrypton/ColabFold), [version](https://colab.research.google.com/github/sokrypton/ColabFold/blob/main/AlphaFold2.ipynb) for homooligomeric prediction and complexes

v t e Google AI Google Google Brain Google DeepMind Computer programs AlphaGo Versions AlphaGo (2015) Master (2016) AlphaGo Zero (2017) AlphaZero (2017) MuZero (2019) Competitions Fan Hui (2015) Lee Sedol (2016) Ke Jie (2017) In popular culture AlphaGo (2017) Other AlphaFold (2018) AlphaStar (2019) AlphaTensor (2022) AlphaDev (2023) AlphaGeometry (2024) AlphaGenome (2025) Machine learning Neural networks Inception (2014) WaveNet (2016) MobileNet (2017) Transformer (2017) EfficientNet (2019) Gato (2022) Other Quantum Artificial Intelligence Lab TensorFlow Tensor Processing Unit Generative AI Chatbots Assistant (2016) Sparrow (2022) Gemini (2023) Nano Banana (2025) Models BERT (2018) XLNet (2019) T5 (2019) LaMDA (2021) Chinchilla (2022) PaLM (2022) Imagen (2023) Gemini (2023) VideoPoet (2024) Gemma (2024) Genie (2024) Veo (2024) Other DreamBooth (2022) NotebookLM (2023) Vids (2024) Gemini Robotics (2025) Antigravity (2025) See also "Attention Is All You Need" Future of Go Summit Generative pre-trained transformer Google Labs Google Workspace Category Commons

v t e Artificial intelligence (AI) History timeline Glossary Companies Projects List of open-source AI software Concepts Automated reasoning Parameter Hyperparameter Loss functions Regression Bias–variance tradeoff Double descent Overfitting Clustering Gradient descent SGD Quasi-Newton method Conjugate gradient method Backpropagation Attention Convolution Normalization Batchnorm Activation Softmax Sigmoid Rectifier Gating Weight initialization Regularization Datasets Augmentation Prompt engineering Reinforcement learning Q-learning SARSA Imitation Policy gradient Diffusion Latent diffusion model Autoregression Adversary RAG Uncanny valley RLHF Self-supervised learning Reflection Recursive self-improvement Hallucination Word embedding Vibe coding Symbolic AI Neuro-symbolic AI Applications Automated theorem proving Machine learning In-context learning Artificial neural network Deep learning Language model Large NMT Reasoning Model Context Protocol Intelligent agent AI agent Artificial human companion Humanity's Last Exam Lethal autonomous weapons (LAWs) Generative AI Weak AI Hypothetical Artificial general intelligence (AGI) Artificial superintelligence (ASI) Agent2Agent protocol Implementations Audio–visual AlexNet WaveNet Human image synthesis HWR OCR Computer vision Speech synthesis 15.ai ElevenLabs Speech recognition Whisper Facial recognition AlphaFold Text-to-image models Aurora DALL-E Firefly Flux GPT Image Ideogram Imagen Midjourney Recraft Stable Diffusion Text-to-video models Dream Machine Runway Gen Hailuo AI Kling Sora Seedance Veo Music generation Riffusion Suno Udio World models Genie Oasis Text List of large language models Project Debater IBM Watson IBM Watsonx Decisional AlphaGo AlphaZero OpenAI Five Self-driving car MuZero Action selection AutoGPT Robot control Reasoning systems Deductive classifiers Expert systems Inference engines Knowledge-based systems Logic programs Procedural reasoning systems Semantic reasoners Rule-based systems Cognitive architectures ACT-R Soar CLARION LIDA OpenCog Knowledge bases ConceptNet Wikidata DBpedia YAGO People Alan Turing Warren Sturgis McCulloch Walter Pitts John von Neumann Christopher D. Manning Claude Shannon Shun'ichi Amari Kunihiko Fukushima Takeo Kanade Marvin Minsky John McCarthy Nathaniel Rochester Allen Newell Cliff Shaw Herbert A. Simon Oliver Selfridge Frank Rosenblatt Bernard Widrow Joseph Weizenbaum Seymour Papert Seppo Linnainmaa Paul Werbos Geoffrey Hinton John Hopfield Jürgen Schmidhuber Yann LeCun Yoshua Bengio Lotfi A. Zadeh Stephen Grossberg Alex Graves James Goodnight Andrew Ng Fei-Fei Li Alex Krizhevsky Ilya Sutskever Oriol Vinyals Quoc V. Le Ian Goodfellow Demis Hassabis David Silver Andrej Karpathy Ashish Vaswani Noam Shazeer Aidan Gomez John Schulman Mustafa Suleyman Jan Leike Daniel Kokotajlo François Chollet Neural network architectures Neural Turing machine Differentiable neural computer Transformer Vision transformer (ViT) Recurrent neural network (RNN) Long short-term memory (LSTM) Gated recurrent unit (GRU) Echo state network Multilayer perceptron (MLP) Convolutional neural network (CNN) Residual neural network (RNN) Highway network Mamba Autoencoder Variational autoencoder (VAE) Generative adversarial network (GAN) Graph neural network (GNN) Political AI Cold War AI safety (Alignment) AI takeover Elections Ethics of AI EU AI Act Nationalism Precautionary principle Regulation of AI US Virtual politician Propaganda Social and economic AI boom AI bubble AI data center AI effect AI literacy AI slop AI winter Anthropomorphism Arms race Competition Environmental impact Explainable AI Generative engine optimization In architecture In education In fiction In healthcare Chatbot psychosis In marketing In video games In visual art Military applications AI warfare Workplace impact Category

---
Adapted from the Wikipedia article [AlphaFold](https://en.wikipedia.org/wiki/AlphaFold) by Wikipedia contributors ([contributor history](https://en.wikipedia.org/wiki/AlphaFold?action=history)). Available under [Creative Commons Attribution-ShareAlike 4.0 International](https://creativecommons.org/licenses/by-sa/4.0/). Changes may have been made.