692:(these relationships are represented by the array shown in red). 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, learnt 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. 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."
47:
818:. 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.
517:
650:
5163:
4239:
951:, the story was widely covered by major national newspapers,. A frequent theme was that ability to predict protein structures accurately based on the constituent amino acid sequence is expected to have a wide variety of benefits in the life sciences space including accelerating advanced drug discovery and enabling better understanding of diseases. Some have noted that even a perfect answer to the protein
5143:
1233:, an international open-access database, before releasing the computationally determined structures of the under-studied protein molecules. 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. Specifically, AlphaFold 2's prediction of the structure of the
766:
4249:
696:
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.
642:
1685:
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).
833:(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, a level of accuracy reported to be comparable to experimental techniques like
444:
were available from proteins with a partially similar sequence. A team that used AlphaFold 2 (2020) repeated the placement in the CASP14 competition in
November 2020. The team achieved a level of accuracy much higher than any other group. It scored above 90 for around two-thirds of the proteins in
856:
To additionally verify AlphaFold-2 the conference organisers approached four leading experimental groups for structures they were finding particularly challenging and had been unable to determine. In all four cases the three-dimensional models produced by AlphaFold 2 were sufficiently accurate to
695:
The output of these iterations then informs the final structure prediction module, which also uses transformers, 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
780:
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
1684:
Jumper, John; Evans, Richard; Pritzel, Alexander; Green, Tim; Figurnov, Michael; Ronneberger, Olaf; Tunyasuvunakool, Kathryn; Bates, Russ; Ĺ˝Ădek, Augustin; Potapenko, Anna; Bridgland, Alex; Meyer, Clemens; Kohl, Simon A A; Ballard, Andrew J; Cowie, Andrew; Romera-Paredes, Bernardino; Nikolov,
1143:, but for humans they are available in the whole batch file. AlphaFold planned to add more sequences to the collection, the initial goal (as of beginning of 2022) being to cover most of the UniRef90 set of more than 100 million proteins. As of May 15, 2022, 992,316 predictions were available.
848:
for the set of overlapped C-alpha atoms. 76% of predictions achieved better than 3 Ă…, and 46% had a C-alpha atom RMS accuracy better than 2 Ă…, with a median RMS deviation in its predictions of 2.1 Ă… for a set of overlapped CA atoms. AlphaFold 2 also achieved an accuracy in modelling surface
699:
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.
664:
The software design used in AlphaFold 1 contained a number of modules, each trained separately, that were used to produce the guide potential that was then combined with the physics-based energy potential. AlphaFold 2 replaced this with a system of sub-networks coupled together into a single
2243:
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).
837:. In 2018 AlphaFold 1 had only reached this level of accuracy in two of all of its predictions. 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.
2660:
For the GDT_TS measure used, each atom in the prediction scores a quarter of a point if it is within 8 Ă… (0.80 nm) of the experimental position; half a point if it is within 4 Ă…, three-quarters of a point if it is within 2 Ă…, and a whole point if it is within 1
923:
called the result "a stunning advance on the protein folding problem", 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."
771:
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
990:
The open access to source code of several AlphaFold versions (excluding AlphaFold 3) has been provided by DeepMind after requests from the scientific community. Full source code of AlphaFold-3 is expected to be provided to open access by the end of 2024.
1203:
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
618:(2018) was built on work developed by various teams in the 2010s, work that looked at the large databanks of related DNA sequences now available from many different organisms (most without known 3D structures), to try to find changes at different
558:, which are all expensive and time-consuming. 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.
749:
AlphaFold 3 introduces the "Pairformer", a deep learning architecture inspired from the transformer, considered similar but simpler than the
Evoformer introduced with AlphaFold 2. The raw predictions from the Pairformer module are passed to a
452:
AlphaFold 2's results at CASP14 were described as "astounding" and "transformational". Some researchers noted that the accuracy is not high enough for a third of its predictions, and that it does not reveal the mechanism or rules of
449:(GDT), a test that measures the degree to which a computational program predicted structure is similar to the lab experiment determined structure, with 100 being a complete match, within the distance cutoff used for calculating GDT.
622:
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
844:(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
884:, a situation AlphaFold was not programmed to consider. 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.
1247:. 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.
793:
807:(GDT) score, ahead of 52.5 and 52.4 by the two next best-placed teams, who were also using deep learning to estimate contact distances. Overall, across all targets, the program achieved a GDT score of 68.5.
783:
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.
665:
differentiable end-to-end model, based entirely on pattern recognition, which was trained in an integrated way as a single integrated structure. Local physics, in the form of energy refinement based on the
1182:
The model relies to some degree upon co-evolutionary information across similar proteins, and thus may not perform well on synthetic proteins or proteins with very low homology to anything in the database.
769:
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.
1200:. Between 50% and 70% of the structures of the human proteome are incomplete without covalently-attached glycans. AlphaFill, a derived database, adds cofactors to AlphaFold models where appropriate.
440:(CASP) in December 2018. The program was particularly successful at predicting the most accurate structure for targets rated as the most difficult by the competition organisers, where no existing
803:
were available from proteins with a partially similar sequence. AlphaFold gave the best prediction for 25 out of 43 protein targets in this class, achieving a median score of 58.9 on the CASP's
3688:
5037:
688:
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
631:
close the residues might be likely to be—turning the contact map into a likely distance map. It also used more advanced learning methods than previously to develop the inference.
306:
869:
1153:
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.
4879:
4285:
3010:
927:
Propelled by press releases from CASP and DeepMind, AlphaFold 2's success received wide media attention. As well as news pieces in the specialist science press, such as
799:
The program was particularly successfully predicting the most accurate structure for targets rated as the most difficult by the competition organisers, where no existing
397:
2984:
1225:
in the United
Kingdom before release into the larger research community. The team also confirmed accurate prediction against the experimentally determined SARS-CoV-2
601:
identify parts of a larger problem, then piece it together to obtain the overall solution. The overall training was conducted on processing power between 100 and 200
2745:
2926:
1214:
979:
841:
2670:
To achieve a GDT_TS score of 92.5, mathematically at least 70% of the structure must be accurate to within 1 Ă…, and at least 85% must be accurate to within 2 Ă…,
589:
DeepMind is known to have trained the program on over 170,000 proteins from a public repository of protein sequences and structures. The program uses a form of
461:
to be considered solved. Nevertheless, there has been widespread respect for the technical achievement. On 15 July 2021 the AlphaFold 2 paper was published in
5212:
437:
158:
2908:
2636:
1437:
2799:
2330:
1997:
669:
model, is applied only as a final refinement step once the neural network prediction has converged, and only slightly adjusts the predicted structure.
2556:
3635:
2762:
436:
AlphaFold software has had three major versions. A team of researchers that used AlphaFold 1 (2018) placed first in the overall rankings of the 13th
4040:
3144:
2780:
257:
235:
4395:
573:
scores of only about 40 out of 100 can be achieved for the most difficult proteins by 2016. AlphaFold started competing in the 2018 CASP using an
193:
171:
2850:
4278:
2891:
2869:
2119:
826:
In
November 2020, DeepMind's new version, AlphaFold 2, won CASP14. Overall, AlphaFold 2 made the best prediction for 88 out of the 97 targets.
95:
5217:
390:
316:
270:
225:
220:
2951:
5068:
1486:
1262:"Protein structure prediction using multiple deep neural networks in the 13th Critical Assessment of Protein Structure Prediction (CASP13)"
754:, which starts with a cloud of atoms and uses these predictions to iteratively progress towards a 3D depiction of the molecular structure.
673:
511:
3695:
1462:
5169:
4720:
4457:
4063:
3827:
975:
369:
341:
336:
230:
4209:
2474:
1240:
329:
198:
188:
178:
2610:
1221:. The structures of these proteins were pending experimental detection in early 2020. Results were examined by the scientists at the
5202:
4981:
4608:
4415:
4271:
3733:
301:
247:
213:
80:
3702:
627:
to be estimated. Building on recent work prior to 2018, AlphaFold 1 extended this to estimate a probability distribution for just
569:, which was launched in 1994 to challenge the scientific community to produce their best protein structure predictions, found that
3284:
McBride, John M.; Polev, Konstantin; Abdirasulov, Amirbek; Reinharz, Vladimir; Grzybowski, Bartosz A.; Tlusty, Tsvi (2023-11-20).
1544:
5197:
4936:
1173:
383:
287:
133:
2140:
2088:
2062:
1326:
1197:
735:
65:
3349:
3064:
5123:
5063:
4661:
3689:
CASP14: what Google DeepMind's AlphaFold 2 really achieved, and what it means for protein folding, biology and bioinformatics
1855:
1598:
Callaway, Ewen (2020-11-30). "'It will change everything': DeepMind's AI makes gigantic leap in solving protein structures".
1360:
1649:
3600:
46:
4656:
4345:
3799:
3756:
Mirdita, Milot; SchĂĽtze, Konstantin; Moriwaki, Yoshitaka; Heo, Lim; Ovchinnikov, Sergey; Steinegger, Martin (2022-05-30).
661:, 2020) is significantly different from the original version that won CASP 13 in 2018, according to the team at DeepMind.
3168:
1663:
5098:
4452:
4405:
4400:
1125:
953:
800:
590:
562:
507:
441:
426:
5149:
4445:
4371:
4141:
681:
252:
203:
100:
1875:
1797:
2211:
2032:
4773:
4708:
4309:
963:
problem—understanding in detail how the folding process actually occurs in nature (and how sometimes they can also
555:
540:
75:
1139:, amounting to over 365,000 proteins. The database does not include proteins with fewer than 16 or more than 2700
5174:
5032:
4671:
4502:
4325:
4198:
2831:
1226:
892:
In 2022 DeepMind did not enter CASP15, but most of the entrants used AlphaFold or tools incorporating AlphaFold.
565:
from their amino acid sequences, but the accuracy of such methods has not been close to experimental techniques.
532:
153:
1827:
5222:
5207:
5073:
4330:
4242:
3934:
2749:
1244:
913:
551:
277:
872:
methods, which define protein structure directly in aqueous solution, whereas AlphaFold was mostly trained on
2581:
774:(Qualitative improvement had been made in earlier years, but it is only as changes bring structures within 8
5118:
5103:
4756:
4751:
4651:
4519:
4300:
4153:
3820:
1380:
1222:
1193:
574:
458:
414:
58:
38:
5078:
4838:
4557:
4552:
4073:
3928:
2729:
Artificial intelligence solution to a 50-year-old science challenge could 'revolutionise' medical research
649:
148:
3551:"AlphaFold Blindness to Topological Barriers Affects Its Ability to Correctly Predict Proteins' Topology"
1851:
5108:
5093:
5058:
4746:
4646:
4514:
4129:
3940:
941:
909:
873:
858:
834:
547:
491:
4976:
3011:"If Google's Alphafold2 really has solved the protein folding problem, they need to show their working"
2488:"Protein tertiary structure modeling driven by deep learning and contact distance prediction in CASP13"
868:
Of the three structures that AlphaFold 2 had the least success in predicting, two had been obtained by
543:
of the proteins. The 3-D structure is crucial to understanding the biological function of the protein.
3382:
5128:
5083:
4529:
4474:
4320:
4315:
4252:
3994:
3307:
3199:
2499:
2166:
1942:
1698:
1607:
905:
830:
811:
804:
570:
468:
446:
90:
3660:
2728:
1192:
AlphaFold 3 version can predict structures of protein complexes with a very limited set of selected
757:
The AlphaFold server was created to provide free access to AlphaFold 3 for non-commercial research.
4703:
4681:
4430:
4425:
4383:
4335:
4204:
2290:
1140:
619:
242:
5142:
5088:
4666:
4495:
3813:
3423:
3297:
3225:
2711:
2419:
2192:
2037:
1631:
1549:
1504:
920:
917:
689:
685:
292:
3498:
Hekkelman, Maarten L.; de Vries, Ida; Joosten, Robbie P.; Perrakis, Anastassis (February 2023).
2083:
The structure module is stated to use a "3-d equivariant transformer architecture" (John Jumper
722:. AlphaFold 3 is not limited to single-chain proteins, as it can also predict the structures of
3661:"How DeepMind's new protein-folding A.I. is already helping to combat the coronavirus pandemic"
3037:
2827:
1314:
1179:
Aphafold-2 was validated for predicting structural effects of mutations with a limited success.
5154:
4946:
4598:
4469:
4462:
3893:
3789:
3730:
Open access to protein structure predictions for the human proteome and 20 other key organisms
3608:
3582:
3531:
3480:
3462:
3415:
3407:
3357:
3331:
3323:
3267:
3217:
2959:
2870:'Once in a generation advance' as Google AI researchers crack 50-year-old biological challenge
2703:
2535:
2517:
2437:
2411:
2403:
2263:
2219:
2184:
2005:
1976:
1958:
1726:
1623:
1574:
1291:
1230:
1218:
1166:
935:
479:
70:
1315:
Fourteenth
Critical Assessment of Techniques for Protein Structure Prediction (Abstract Book)
1261:
4899:
4889:
4696:
4490:
4440:
4435:
4378:
4366:
4224:
4112:
4100:
3779:
3769:
3572:
3562:
3521:
3511:
3470:
3454:
3397:
3315:
3207:
2815:
2693:
2525:
2509:
2466:
2393:
2271:
2253:
2174:
1966:
1950:
1894:
1886:
1867:
1716:
1706:
1615:
1494:
1299:
1272:
1165:
AlphaFold DB provides monomeric models of proteins, rather than their biologically relevant
1095:
929:
862:
463:
208:
143:
128:
5012:
4956:
4778:
4420:
4340:
4046:
3857:
3383:"The case for post-predictional modifications in the AlphaFold Protein Structure Database"
3285:
3271:
2985:"La inteligencia artificial arrasa en uno de los problemas más importantes de la biologĂa"
2874:
2098:
2094:
1136:
1080:
1035:
959:
751:
723:
715:
536:
454:
85:
3577:
3550:
3311:
3203:
2313:
2276:
2170:
1946:
1702:
1611:
792:
In
December 2018, DeepMind's AlphaFold placed first in the overall rankings of the 13th
4986:
4951:
4941:
4766:
4524:
4350:
3958:
3784:
3757:
3526:
3499:
3475:
3443:"Determination of glycosylation sites and site-specific heterogeneity in glycoproteins"
3442:
3033:
2836:
2530:
2487:
1971:
1930:
1721:
1686:
1401:
971:
881:
877:
677:
516:
1312:(December 2020), "High Accuracy Protein Structure Prediction Using Deep Learning", in
5191:
4931:
4911:
4828:
4507:
4106:
4006:
3427:
3263:
3229:
3053:
2785:
2767:
2715:
2423:
2355:
2245:
2196:
1881:
1635:
1508:
1375:
1345:
947:
719:
594:
578:
430:
422:
138:
2927:"2023 Breakthrough Prizes Announced: Deepmind's Protein Folders Awarded $ 3 Million"
2116:. It is not known how similar this may or may not be to what was used in AlphaFold.
1871:
1286:
5017:
4848:
4219:
3899:
3852:
3319:
2912:
2855:
2637:"DeepMind's protein-folding AI has solved a 50-year-old grand challenge of biology"
1438:"DeepMind's protein-folding AI has solved a 50-year-old grand challenge of biology"
1355:
1340:
1186:
810:
In
January 2020, implementations and illustrative code of AlphaFold 1 was released
778:
of their experimental positions that they start to affect the CASP GDS-TS measure).
282:
2113:
5113:
4884:
4793:
4788:
4410:
4388:
4214:
4176:
3739:
3188:"'The entire protein universe': AI predicts shape of nearly every known protein"
2909:
Protein folding and related problems remain unsolved despite AlphaFold's advance
2887:
1798:"AlphaFold 3 predicts the structure and interactions of all of life's molecules"
964:
765:
624:
521:
311:
296:
3774:
3516:
3458:
3402:
3212:
3187:
2698:
2681:
2258:
2179:
2154:
1954:
1898:
1711:
1619:
546:
Protein structures can be determined experimentally through techniques such as
5007:
4966:
4961:
4874:
4783:
4691:
4603:
4583:
4068:
3966:
3745:
3567:
3381:
Bagdonas, Haroldas; Fogarty, Carl A.; Fadda, Elisa; Agirre, Jon (2021-10-29).
2895:
2398:
2381:
2123:
1499:
1303:
850:
3612:
3466:
3411:
3361:
3327:
2963:
2521:
2407:
2267:
2246:"Accurate structure prediction of biomolecular interactions with AlphaFold 3"
2223:
2009:
1962:
5002:
4971:
4869:
4713:
4613:
4567:
4562:
4547:
4159:
3905:
3836:
3246:
2804:
2454:
1365:
1172:
Many protein regions are predicted with low confidence score, including the
1100:
346:
110:
4263:
3793:
3586:
3535:
3484:
3419:
3335:
3221:
2707:
2539:
2415:
2291:
Accurate structure prediction of biomolecular interactions with AlphaFold 3
2188:
1980:
1730:
1627:
1287:"Improved protein structure prediction using potentials from deep learning"
561:
Over the years, researchers have applied numerous computational methods to
17:
3636:"Computational predictions of protein structures associated with COVID-19"
4904:
4736:
4000:
3751:
3715:
2302:
1463:"DeepMind solves 50-year-old 'grand challenge' with protein folding A.I."
1147:
1132:
1121:
1020:
845:
775:
672:
A key part of the 2020 system are two modules, believed to be based on a
478:
AlphaFold 3 was announced on 8 May 2024. It can predict the structure of
472:
418:
183:
105:
3720:
2892:
The secret of life, part 2: the solution of the protein folding problem.
2832:
London A.I. Lab Claims
Breakthrough That Could Accelerate Drug Discovery
2810:
In all science editor Tom
Whipple wrote six articles on the subject for
1744:
641:
5027:
4864:
4818:
4741:
4641:
4636:
4588:
4034:
3887:
3100:
3041:
2513:
2470:
2455:
Deep-learning contact-map guided protein structure prediction in CASP13
2109:
2105:
1370:
1276:
1239:
protein was very similar to the structure determined by researchers at
1129:
1120:
was launched on July 22, 2021, as a joint effort between AlphaFold and
528:
351:
2682:"After AlphaFold: protein-folding contest seeks next big breakthrough"
2106:
SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks
5042:
5022:
4894:
4686:
3911:
3847:
3724:
2763:
DeepMind's AI biologist can decipher secrets of the machinery of life
1913:
1909:
1350:
912:
and great progress towards a decades-old grand challenge of biology.
815:
739:
3350:"DeepMind's latest AI breakthrough could turbocharge drug discovery"
2800:
Deepmind finds biology's 'holy grail' with answer to protein problem
2781:
The predictions of DeepMind's latest AI could revolutionise medicine
2746:
Protein folding and science communication: Between hype and humility
3441:
An, Hyun Joo; Froehlich, John W; Lebrilla, Carlito B (2009-10-01).
3302:
2611:"AlphaFold: a solution to a 50-year-old grand challenge in biology"
2504:
1060:
4843:
4823:
4813:
4808:
4803:
4798:
4761:
4593:
4135:
3500:"AlphaFill: enriching AlphaFold models with ligands and cofactors"
3264:
AlphaFold heralds a data-driven revolution in biology and medicine
3125:
3076:
1235:
794:
Critical Assessment of Techniques for Protein Structure Prediction
764:
666:
648:
640:
2155:"What's next for AlphaFold and the AI protein-folding revolution"
1545:'The game has changed.' AI triumphs at solving protein structures
4833:
4147:
2732:
2557:"DeepMind Breakthrough Helps to Solve How Diseases Invade Cells"
2486:
Hou, Jie; Wu, Tianqi; Cao, Renzhi; Cheng, Jianlin (2019-04-25).
2212:"Google Unveils A.I. for Predicting Behavior of Human Molecules"
901:
602:
566:
4267:
3809:
3729:
3696:
AlphaFold2 @ CASP14: "It feels like one's child has left home."
3549:
Dabrowski-Tumanski, Pawel; Stasiak, Andrzej (7 November 2023).
1768:
1150:
has been updated to show AlphaFold predictions when available.
743:
731:
727:
495:
487:
483:
3065:
AlphaFold 3: Stepping into the future of structure prediction
1931:"Highly accurate protein structure prediction with AlphaFold"
1745:"GitHub - deepmind/alphafold: Open source code for AlphaFold"
1687:"Highly accurate protein structure prediction with AlphaFold"
1213:
AlphaFold has been used to predict structures of proteins of
3805:
2033:
Lessons from DeepMind's breakthrough in protein-folding A.I.
1910:
AlphaFold: Machine learning for protein structure prediction
45:
2952:"Team behind AI program AlphaFold win Lasker science prize"
1998:"DeepMind is answering one of biology's biggest challenges"
598:
2851:
DeepMind AI cracks 50-year-old problem of protein folding
3054:
AlphaFold3 — why did Nature publish it without its code?
3169:"Putting the power of AlphaFold into the world's hands"
3145:"Alphafold Structure Predictions Available In Interpro"
3038:"Brief update on some exciting progress on #AlphaFold!"
2074:
John Jumper et al., conference abstract (December 2020)
1124:. At launch the database contains AlphaFold-predicted
957:
problem would still leave questions about the protein
645:
AlphaFold 2 performance, experiments, and architecture
3758:"ColabFold: Making protein folding accessible to all"
2316:, CASP 13, December 2018. (AlphaFold = Team 043: A7D)
1487:"Structural biology: How proteins got their close-up"
1992:
1990:
5051:
4995:
4924:
4857:
4729:
4629:
4622:
4576:
4540:
4483:
4359:
4299:
4190:
4169:
4122:
4093:
4086:
4056:
4027:
4020:
3981:
3950:
3921:
3880:
3873:
3866:
1882:
A watershed moment for protein structure prediction
1874:. See also Mohammed AlQuraishi (December 9, 2018),
1106:
1094:
1089:
1079:
1074:
1066:
1056:
1051:
1044:
1034:
1029:
1019:
1009:
1004:
3746:AlphaFold: The making of a scientific breakthrough
3703:The AlphaFold2 Method Paper: A Fount of Good Ideas
2331:"Google's DeepMind predicts 3D shapes of proteins"
1575:CASP14 scores just came out and they're astounding
2605:
2603:
2601:
2492:Proteins: Structure, Function, and Bioinformatics
2459:Proteins: Structure, Function, and Bioinformatics
1679:
1677:
1569:
1567:
1565:
1563:
1561:
1559:
1539:
1537:
1535:
1533:
1531:
1529:
1527:
1525:
995:Database of protein models generated by AlphaFold
908:(GDT) is considered a significant achievement in
861:. These included target T1100 (Af1503), a small
676:design, which are used to progressively refine a
3691:, Oxford Protein Informatics Group. (3 December)
3601:"AI Can Help Scientists Find a Covid-19 Vaccine"
3286:"AlphaFold2 Can Predict Single-Mutation Effects"
3241:
3239:
2053:
2051:
2049:
2047:
2324:
2322:
1432:
1430:
1428:
1426:
1424:
1422:
982:for their management of the AlphaFold project.
781:surpassed this again in 2020, across the board.
2356:"AlphaFold: Using AI for scientific discovery"
2027:
2025:
1828:"AlphaFold: Using AI for scientific discovery"
980:Albert Lasker Award for Basic Medical Research
880:consisting of 52 identical copies of the same
4279:
3821:
3274:, volume 12, pages 1666–1669, 12 October 2021
1266:Proteins: Structure, Function, Bioinformatics
1185:The ability of the model to produce multiple
391:
8:
3270:, Roman A. Laskowski & Neera Borkakoti,
3247:"What use cases does AlphaFold not support?"
2314:Group performance based on combined z-scores
999:
853:described as "really really extraordinary".
3802:for homooligomeric prediction and complexes
2293:, pdf of preprint of the article in Nature.
1650:No, DeepMind has not solved protein folding
865:studied by experimentalists for ten years.
467:as an advance access publication alongside
438:Critical Assessment of Structure Prediction
27:Artificial intelligence program by DeepMind
4626:
4286:
4272:
4264:
4090:
4024:
3877:
3870:
3828:
3814:
3806:
2093:One design for a transformer network with
1652:, Reciprocal Space (blog), 2 December 2020
998:
857:determine structures of these proteins by
684:" in graph-theory terminology) between an
398:
384:
29:
3783:
3773:
3576:
3566:
3525:
3515:
3474:
3401:
3390:Nature Structural & Molecular Biology
3301:
3211:
2697:
2529:
2503:
2397:
2275:
2257:
2178:
1970:
1929:Jumper, John; et al. (August 2021).
1924:
1922:
1876:AlphaFold @ CASP13: "What just happened?"
1720:
1710:
1498:
1128:of protein structures of nearly the full
2440:for 043 A7D, 322 Zhang, and 089 MULTICOM
714:was co-developed by Google DeepMind and
515:
1880:Mohammed AlQuraishi (15 January 2020),
1393:
37:
3736:(AlphaFold Protein Structure Database)
3721:AlphaFold v2.1 code and links to model
3126:"AlphaFold Protein Structure Database"
3101:"AlphaFold Protein Structure Database"
3077:"AlphaFold Protein Structure Database"
2735:organising committee, 30 November 2020
2303:A non-commercial server of AlphaFold-3
1769:"AlphaFold Protein Structure Database"
3630:
3628:
2551:
2549:
2057:See block diagram. Also John Jumper
1822:
1820:
1818:
1792:
1790:
1788:
1189:conformations of proteins is limited.
597:technique that focuses on having the
563:predict the 3D structures of proteins
471:and a searchable database of species
7:
5124:Generative adversarial network (GAN)
4248:
3449:. Analytical Techniques/Mechanisms.
3251:AlphaFold Protein Structure Database
1593:
1591:
1589:
1587:
1585:
1583:
1485:Stoddart, Charlotte (1 March 2022).
1118:AlphaFold Protein Structure Database
1000:AlphaFold Protein Structure Database
900:AlphaFold 2 scoring more than 90 in
653:Architectural details of AlphaFold 2
512:De novo protein structure prediction
5213:Deep learning software applications
4064:Quantum Artificial Intelligence Lab
3447:Current Opinion in Chemical Biology
1161:AlphaFold has various limitations:
976:Breakthrough Prize in Life Sciences
876:. The third exists in nature as a
4210:Generative pre-trained transformer
1662:Balls, Phillip (9 December 2020).
1241:University of California, Berkeley
541:three dimensional (3-D) structures
25:
3734:European Bioinformatics Institute
1664:"Behind the screens of AlphaFold"
1329:". Presentation given at CASP 14.
657:The 2020 version of the program (
5162:
5161:
5141:
4247:
4238:
4237:
2153:Callaway, Ewen (13 April 2022).
1198:post-translational modifications
1174:intrinsically disordered protein
1061:https://www.alphafold.ebi.ac.uk/
736:post-translational modifications
427:predictions of protein structure
2329:Sample, Ian (2 December 2018).
2122:by AlQuaraishi on this, or the
829:On the competition's preferred
429:. The program is designed as a
66:Artificial general intelligence
5074:Recurrent neural network (RNN)
5064:Differentiable neural computer
3320:10.1103/PhysRevLett.131.218401
2983:DomĂnguez, Nuño (2020-12-02).
1361:Human Proteome Folding Project
874:protein structures in crystals
1:
5119:Variational autoencoder (VAE)
5079:Long short-term memory (LSTM)
4346:Computational learning theory
3186:Callaway, Ewen (2022-07-28).
2680:Callaway, Ewen (2022-12-13).
2382:"Deep learning 3D structures"
2101:was proposed in Fabian Fuchs
1872:10.1093/bioinformatics/btz422
1146:In July 2021, UniProt-KB and
5218:Molecular modelling software
5099:Convolutional neural network
3009:Briggs, David (2020-12-04).
2814:on the day the news broke. (
2582:"deepmind/deepmind-research"
1577:, Twitter, 30 November 2020.
1015:protein structure prediction
970:In 2023, Demis Hassabis and
520:Amino-acid chains, known as
508:Protein structure prediction
5094:Multilayer perceptron (MLP)
2808:(online), 30 November 2020.
680:for each relationship (or "
101:Natural language processing
5239:
5170:Artificial neural networks
5084:Gated recurrent unit (GRU)
4310:Differentiable programming
3775:10.1038/s41592-022-01488-1
3517:10.1038/s41592-022-01685-y
3459:10.1016/j.cbpa.2009.07.022
3403:10.1038/s41594-021-00680-9
3213:10.1038/d41586-022-02083-2
2950:Sample, Ian (2023-09-21).
2699:10.1038/d41586-022-04438-1
2259:10.1038/s41586-024-07487-w
2180:10.1038/d41586-022-00997-5
1955:10.1038/s41586-021-03819-2
1899:10.1038/d41586-019-03951-0
1712:10.1038/s41586-021-03819-2
1620:10.1038/d41586-020-03348-4
842:root-mean-square deviation
556:nuclear magnetic resonance
505:
417:(AI) program developed by
154:Hybrid intelligent systems
76:Recursive self-improvement
5137:
4503:Artificial neural network
4326:Automatic differentiation
4233:
4199:Attention Is All You Need
3843:
3568:10.3390/molecules28227462
3149:proteinswebteam.github.io
3143:InterPro (22 July 2021).
2399:10.1038/s41592-020-0779-y
2210:Metz, Cade (2024-05-08).
1500:10.1146/knowable-022822-1
1461:Shead, Sam (2020-11-30).
1304:10.1038/s41586-019-1923-7
1217:, the causative agent of
710:Announced on 8 May 2024,
524:, fold to form a protein.
482:created by proteins with
5203:Applied machine learning
4331:Neuromorphic engineering
4294:Differentiable computing
3748:, DeepMind, via YouTube.
2750:University of Nottingham
2498:(12). Wiley: 1165–1178.
2141:AlphaFold 2 presentation
2089:AlphaFold 2 presentation
2063:AlphaFold 2 presentation
1245:cryo-electron microscopy
552:cryo-electron microscopy
278:Artificial consciousness
5198:Bioinformatics software
5104:Residual neural network
4520:Artificial Intelligence
3698:(blog), 8 December 2020
3290:Physical Review Letters
3130:www.alphafold.ebi.ac.uk
2380:Singh, Arunima (2020).
1381:Predicted Aligned Error
1229:that was shared in the
1223:Francis Crick Institute
718:, both subsidiaries of
575:artificial intelligence
459:protein folding problem
415:artificial intelligence
149:Evolutionary algorithms
39:Artificial intelligence
4074:Tensor Processing Unit
3716:AlphaFold-3 web server
784:
654:
646:
525:
50:
5059:Neural Turing machine
4647:Human image synthesis
3701:Mohammed AlQuraishi,
3694:Mohammed AlQuraishi,
2915:blog, 8 December 2020
2752:blog, 4 December 2020
2641:MIT Technology Review
1573:Mohammed AlQuraishi,
1442:MIT Technology Review
1025:all UniProt proteomes
942:MIT Technology Review
910:computational biology
859:molecular replacement
835:X-ray crystallography
768:
678:vector of information
652:
644:
548:X-ray crystallography
533:chains of amino acids
519:
49:
5150:Computer programming
5129:Graph neural network
4704:Text-to-video models
4682:Text-to-image models
4530:Large language model
4515:Scientific computing
4321:Statistical manifold
4316:Information geometry
3705:(blog), 25 July 2021
918:structural biologist
906:global distance test
831:global distance test
805:global distance test
469:open source software
447:global distance test
91:General game playing
4496:In-context learning
4336:Pattern recognition
4205:Future of Go Summit
3312:2023PhRvL.131u8401M
3204:2022Natur.608...15C
3105:alphafold.ebi.ac.uk
3081:alphafold.ebi.ac.uk
2438:CASP 13 data tables
2171:2022Natur.604..234C
2139:(1 December 2020),
2087:(1 December 2020),
2061:(1 December 2020),
1947:2021Natur.596..583J
1856:AlphaFold at CASP13
1852:Mohammed AlQuraishi
1773:alphafold.ebi.ac.uk
1703:2021Natur.596..583J
1612:2020Natur.588..203C
1543:Robert F. Service,
1285:(15 January 2020),
1141:amino acid residues
1001:
878:multidomain complex
801:template structures
442:template structures
243:Machine translation
159:Systems integration
96:Knowledge reasoning
33:Part of a series on
5089:Echo state network
4977:JĂĽrgen Schmidhuber
4672:Facial recognition
4667:Speech recognition
4577:Software libraries
3951:In popular culture
2907:e.g. Greg Bowman,
2898:, 30 November 2020
2878:, 30 November 2020
2859:, 30 November 2020
2840:, 30 November 2020
2771:, 30 November 2020
2744:Brigitte Nerlich,
2617:. 30 November 2020
2514:10.1002/prot.25697
2471:10.1002/prot.25792
2216:The New York Times
2124:more detailed post
1553:, 30 November 2020
1325:(December 2020), "
1277:10.1002/prot.25834
921:Venki Ramakrishnan
785:
704:AlphaFold 3 (2024)
690:sequence alignment
686:amino acid residue
655:
647:
635:AlphaFold 2 (2020)
609:AlphaFold 1 (2018)
537:spontaneously fold
526:
421:, a subsidiary of
51:
5185:
5184:
4947:Stephen Grossberg
4920:
4919:
4261:
4260:
4186:
4185:
4082:
4081:
4016:
4015:
3977:
3976:
3867:Computer programs
3687:Carlos Outeiral,
3268:Janet M. Thornton
2789:, 2 December 2020
2761:Michael Le Page,
2731:(press release),
2165:(7905): 234–238.
2143:, slides 12 to 20
2041:, 1 December 2020
1941:(7873): 583–589.
1916:, 31 January 2020
1834:. 15 January 2020
1697:(7873): 583–589.
1606:(7837): 203–204.
1491:Knowable Magazine
1281:Andrew W. Senior
1260:(December 2019),
1256:Andrew W. Senior
1231:Protein Data Bank
1135:of humans and 20
1114:
1113:
724:protein complexes
591:attention network
425:, which performs
408:
407:
144:Bayesian networks
71:Intelligent agent
16:(Redirected from
5230:
5175:Machine learning
5165:
5164:
5145:
4900:Action selection
4890:Self-driving car
4697:Stable Diffusion
4662:Speech synthesis
4627:
4491:Machine learning
4367:Gradient descent
4288:
4281:
4274:
4265:
4251:
4250:
4241:
4240:
4225:Google Workspace
4091:
4025:
4021:Machine learning
3878:
3871:
3830:
3823:
3816:
3807:
3797:
3787:
3777:
3675:
3674:
3672:
3671:
3657:
3651:
3650:
3648:
3647:
3632:
3623:
3622:
3620:
3619:
3597:
3591:
3590:
3580:
3570:
3546:
3540:
3539:
3529:
3519:
3495:
3489:
3488:
3478:
3438:
3432:
3431:
3405:
3387:
3378:
3372:
3371:
3369:
3368:
3346:
3340:
3339:
3305:
3281:
3275:
3261:
3255:
3254:
3243:
3234:
3233:
3215:
3183:
3177:
3176:
3165:
3159:
3158:
3156:
3155:
3140:
3134:
3133:
3122:
3116:
3115:
3113:
3111:
3097:
3091:
3090:
3088:
3087:
3073:
3067:
3062:
3056:
3051:
3045:
3031:
3025:
3024:
3022:
3021:
3006:
3000:
2999:
2997:
2996:
2980:
2974:
2973:
2971:
2970:
2947:
2941:
2940:
2938:
2937:
2922:
2916:
2905:
2899:
2885:
2879:
2868:Lizzie Roberts,
2866:
2860:
2847:
2841:
2825:
2819:
2796:
2790:
2778:
2772:
2759:
2753:
2742:
2736:
2726:
2720:
2719:
2701:
2677:
2671:
2668:
2662:
2658:
2652:
2651:
2649:
2647:
2633:
2627:
2626:
2624:
2622:
2607:
2596:
2595:
2593:
2592:
2578:
2572:
2571:
2569:
2568:
2553:
2544:
2543:
2533:
2507:
2483:
2477:
2447:
2441:
2434:
2428:
2427:
2401:
2377:
2371:
2370:
2368:
2366:
2352:
2346:
2345:
2343:
2341:
2326:
2317:
2311:
2305:
2300:
2294:
2288:
2282:
2281:
2279:
2261:
2240:
2234:
2233:
2231:
2230:
2207:
2201:
2200:
2182:
2150:
2144:
2133:
2127:
2081:
2075:
2072:
2066:
2055:
2042:
2029:
2020:
2019:
2017:
2016:
1994:
1985:
1984:
1974:
1926:
1917:
1907:
1901:
1866:(22), 4862–4865
1849:
1843:
1842:
1840:
1839:
1824:
1813:
1812:
1810:
1809:
1794:
1783:
1782:
1780:
1779:
1765:
1759:
1758:
1756:
1755:
1741:
1735:
1734:
1724:
1714:
1681:
1672:
1671:
1659:
1653:
1646:
1640:
1639:
1595:
1578:
1571:
1554:
1541:
1520:
1519:
1517:
1515:
1502:
1482:
1476:
1475:
1473:
1472:
1458:
1452:
1451:
1449:
1448:
1434:
1417:
1416:
1414:
1412:
1398:
1318:, pp. 22–24
1045:Primary citation
1002:
863:membrane protein
840:Measured by the
400:
393:
386:
307:Existential risk
129:Machine learning
30:
21:
5238:
5237:
5233:
5232:
5231:
5229:
5228:
5227:
5223:Google DeepMind
5208:Protein folding
5188:
5187:
5186:
5181:
5133:
5047:
5013:Google DeepMind
4991:
4957:Geoffrey Hinton
4916:
4853:
4779:Project Debater
4725:
4623:Implementations
4618:
4572:
4536:
4479:
4421:Backpropagation
4355:
4341:Tensor calculus
4295:
4292:
4262:
4257:
4229:
4182:
4165:
4123:Language models
4118:
4078:
4052:
4028:Neural networks
4012:
3973:
3946:
3917:
3862:
3858:Google DeepMind
3839:
3834:
3755:
3712:
3684:
3682:Further reading
3679:
3678:
3669:
3667:
3659:
3658:
3654:
3645:
3643:
3642:. 4 August 2020
3634:
3633:
3626:
3617:
3615:
3599:
3598:
3594:
3548:
3547:
3543:
3497:
3496:
3492:
3440:
3439:
3435:
3396:(11): 869–870.
3385:
3380:
3379:
3375:
3366:
3364:
3348:
3347:
3343:
3283:
3282:
3278:
3272:Nature Medicine
3262:
3258:
3245:
3244:
3237:
3198:(7921): 15–16.
3185:
3184:
3180:
3175:. 22 July 2022.
3167:
3166:
3162:
3153:
3151:
3142:
3141:
3137:
3124:
3123:
3119:
3109:
3107:
3099:
3098:
3094:
3085:
3083:
3075:
3074:
3070:
3063:
3059:
3052:
3048:
3032:
3028:
3019:
3017:
3008:
3007:
3003:
2994:
2992:
2982:
2981:
2977:
2968:
2966:
2949:
2948:
2944:
2935:
2933:
2924:
2923:
2919:
2906:
2902:
2886:
2882:
2875:Daily Telegraph
2867:
2863:
2848:
2844:
2826:
2822:
2809:
2797:
2793:
2779:
2775:
2760:
2756:
2743:
2739:
2727:
2723:
2692:(7942): 13–14.
2679:
2678:
2674:
2669:
2665:
2659:
2655:
2645:
2643:
2635:
2634:
2630:
2620:
2618:
2609:
2608:
2599:
2590:
2588:
2580:
2579:
2575:
2566:
2564:
2555:
2554:
2547:
2485:
2484:
2480:
2465:(12) 1149–1164
2448:
2444:
2435:
2431:
2379:
2378:
2374:
2364:
2362:
2354:
2353:
2349:
2339:
2337:
2328:
2327:
2320:
2312:
2308:
2301:
2297:
2289:
2285:
2242:
2241:
2237:
2228:
2226:
2209:
2208:
2204:
2152:
2151:
2147:
2134:
2130:
2126:by Fabian Fuchs
2117:
2092:
2082:
2078:
2073:
2069:
2056:
2045:
2030:
2023:
2014:
2012:
1996:
1995:
1988:
1928:
1927:
1920:
1908:
1904:
1879:
1850:
1846:
1837:
1835:
1826:
1825:
1816:
1807:
1805:
1796:
1795:
1786:
1777:
1775:
1767:
1766:
1762:
1753:
1751:
1743:
1742:
1738:
1683:
1682:
1675:
1668:Chemistry World
1661:
1660:
1656:
1648:Stephen Curry,
1647:
1643:
1597:
1596:
1581:
1572:
1557:
1542:
1523:
1513:
1511:
1484:
1483:
1479:
1470:
1468:
1460:
1459:
1455:
1446:
1444:
1436:
1435:
1420:
1410:
1408:
1400:
1399:
1395:
1390:
1385:
1336:
1271:(12) 1141–1148
1253:
1251:Published works
1211:
1159:
1137:model organisms
1107:Curation policy
1036:Research center
1011:
997:
988:
978:as well as the
898:
890:
824:
790:
782:
779:
773:
770:
763:
752:diffusion model
716:Isomorphic Labs
706:
637:
611:
587:
514:
504:
455:protein folding
404:
375:
374:
365:
357:
356:
332:
322:
321:
293:Control problem
273:
263:
262:
174:
164:
163:
124:
116:
115:
86:Computer vision
61:
28:
23:
22:
15:
12:
11:
5:
5236:
5234:
5226:
5225:
5220:
5215:
5210:
5205:
5200:
5190:
5189:
5183:
5182:
5180:
5179:
5178:
5177:
5172:
5159:
5158:
5157:
5152:
5138:
5135:
5134:
5132:
5131:
5126:
5121:
5116:
5111:
5106:
5101:
5096:
5091:
5086:
5081:
5076:
5071:
5066:
5061:
5055:
5053:
5049:
5048:
5046:
5045:
5040:
5035:
5030:
5025:
5020:
5015:
5010:
5005:
4999:
4997:
4993:
4992:
4990:
4989:
4987:Ilya Sutskever
4984:
4979:
4974:
4969:
4964:
4959:
4954:
4952:Demis Hassabis
4949:
4944:
4942:Ian Goodfellow
4939:
4934:
4928:
4926:
4922:
4921:
4918:
4917:
4915:
4914:
4909:
4908:
4907:
4897:
4892:
4887:
4882:
4877:
4872:
4867:
4861:
4859:
4855:
4854:
4852:
4851:
4846:
4841:
4836:
4831:
4826:
4821:
4816:
4811:
4806:
4801:
4796:
4791:
4786:
4781:
4776:
4771:
4770:
4769:
4759:
4754:
4749:
4744:
4739:
4733:
4731:
4727:
4726:
4724:
4723:
4718:
4717:
4716:
4711:
4701:
4700:
4699:
4694:
4689:
4679:
4674:
4669:
4664:
4659:
4654:
4649:
4644:
4639:
4633:
4631:
4624:
4620:
4619:
4617:
4616:
4611:
4606:
4601:
4596:
4591:
4586:
4580:
4578:
4574:
4573:
4571:
4570:
4565:
4560:
4555:
4550:
4544:
4542:
4538:
4537:
4535:
4534:
4533:
4532:
4525:Language model
4522:
4517:
4512:
4511:
4510:
4500:
4499:
4498:
4487:
4485:
4481:
4480:
4478:
4477:
4475:Autoregression
4472:
4467:
4466:
4465:
4455:
4453:Regularization
4450:
4449:
4448:
4443:
4438:
4428:
4423:
4418:
4416:Loss functions
4413:
4408:
4403:
4398:
4393:
4392:
4391:
4381:
4376:
4375:
4374:
4363:
4361:
4357:
4356:
4354:
4353:
4351:Inductive bias
4348:
4343:
4338:
4333:
4328:
4323:
4318:
4313:
4305:
4303:
4297:
4296:
4293:
4291:
4290:
4283:
4276:
4268:
4259:
4258:
4256:
4255:
4245:
4234:
4231:
4230:
4228:
4227:
4222:
4217:
4212:
4207:
4202:
4194:
4192:
4188:
4187:
4184:
4183:
4181:
4180:
4173:
4171:
4167:
4166:
4164:
4163:
4157:
4151:
4145:
4139:
4133:
4126:
4124:
4120:
4119:
4117:
4116:
4110:
4104:
4097:
4095:
4088:
4084:
4083:
4080:
4079:
4077:
4076:
4071:
4066:
4060:
4058:
4054:
4053:
4051:
4050:
4044:
4038:
4031:
4029:
4022:
4018:
4017:
4014:
4013:
4011:
4010:
4004:
3998:
3992:
3985:
3983:
3979:
3978:
3975:
3974:
3972:
3971:
3963:
3954:
3952:
3948:
3947:
3945:
3944:
3938:
3932:
3925:
3923:
3919:
3918:
3916:
3915:
3909:
3903:
3897:
3891:
3884:
3882:
3875:
3868:
3864:
3863:
3861:
3860:
3855:
3850:
3844:
3841:
3840:
3835:
3833:
3832:
3825:
3818:
3810:
3804:
3803:
3768:(6): 679–682.
3762:Nature Methods
3749:
3743:
3737:
3727:
3718:
3711:
3710:External links
3708:
3707:
3706:
3699:
3692:
3683:
3680:
3677:
3676:
3652:
3624:
3592:
3541:
3510:(2): 205–213.
3504:Nature Methods
3490:
3453:(4): 421–426.
3433:
3373:
3341:
3296:(21): 218401.
3276:
3256:
3235:
3178:
3160:
3135:
3117:
3092:
3068:
3057:
3046:
3044:, 18 June 2021
3034:Demis Hassabis
3026:
3001:
2975:
2942:
2917:
2900:
2880:
2861:
2842:
2837:New York Times
2820:
2791:
2773:
2754:
2737:
2721:
2672:
2663:
2653:
2628:
2597:
2573:
2545:
2505:10.1101/552422
2478:
2442:
2429:
2386:Nature Methods
2372:
2347:
2318:
2306:
2295:
2283:
2235:
2202:
2145:
2128:
2076:
2067:
2043:
2021:
2004:. 2020-11-30.
1986:
1918:
1902:
1860:Bioinformatics
1844:
1814:
1784:
1760:
1736:
1673:
1654:
1641:
1579:
1555:
1521:
1477:
1453:
1418:
1392:
1391:
1389:
1386:
1384:
1383:
1378:
1373:
1368:
1363:
1358:
1353:
1348:
1343:
1337:
1335:
1332:
1331:
1330:
1319:
1306:
1279:
1252:
1249:
1210:
1207:
1206:
1205:
1201:
1190:
1183:
1180:
1177:
1170:
1158:
1155:
1112:
1111:
1108:
1104:
1103:
1098:
1092:
1091:
1087:
1086:
1083:
1077:
1076:
1072:
1071:
1068:
1064:
1063:
1058:
1054:
1053:
1049:
1048:
1046:
1042:
1041:
1038:
1032:
1031:
1027:
1026:
1023:
1017:
1016:
1013:
1007:
1006:
996:
993:
987:
984:
897:
894:
889:
886:
823:
820:
789:
786:
762:
759:
705:
702:
636:
633:
610:
607:
586:
583:
503:
500:
406:
405:
403:
402:
395:
388:
380:
377:
376:
373:
372:
366:
363:
362:
359:
358:
355:
354:
349:
344:
339:
333:
328:
327:
324:
323:
320:
319:
314:
309:
304:
299:
290:
285:
280:
274:
269:
268:
265:
264:
261:
260:
255:
250:
245:
240:
239:
238:
228:
223:
218:
217:
216:
211:
206:
196:
191:
189:Earth sciences
186:
181:
179:Bioinformatics
175:
170:
169:
166:
165:
162:
161:
156:
151:
146:
141:
136:
131:
125:
122:
121:
118:
117:
114:
113:
108:
103:
98:
93:
88:
83:
78:
73:
68:
62:
57:
56:
53:
52:
42:
41:
35:
34:
26:
24:
14:
13:
10:
9:
6:
4:
3:
2:
5235:
5224:
5221:
5219:
5216:
5214:
5211:
5209:
5206:
5204:
5201:
5199:
5196:
5195:
5193:
5176:
5173:
5171:
5168:
5167:
5160:
5156:
5153:
5151:
5148:
5147:
5144:
5140:
5139:
5136:
5130:
5127:
5125:
5122:
5120:
5117:
5115:
5112:
5110:
5107:
5105:
5102:
5100:
5097:
5095:
5092:
5090:
5087:
5085:
5082:
5080:
5077:
5075:
5072:
5070:
5067:
5065:
5062:
5060:
5057:
5056:
5054:
5052:Architectures
5050:
5044:
5041:
5039:
5036:
5034:
5031:
5029:
5026:
5024:
5021:
5019:
5016:
5014:
5011:
5009:
5006:
5004:
5001:
5000:
4998:
4996:Organizations
4994:
4988:
4985:
4983:
4980:
4978:
4975:
4973:
4970:
4968:
4965:
4963:
4960:
4958:
4955:
4953:
4950:
4948:
4945:
4943:
4940:
4938:
4935:
4933:
4932:Yoshua Bengio
4930:
4929:
4927:
4923:
4913:
4912:Robot control
4910:
4906:
4903:
4902:
4901:
4898:
4896:
4893:
4891:
4888:
4886:
4883:
4881:
4878:
4876:
4873:
4871:
4868:
4866:
4863:
4862:
4860:
4856:
4850:
4847:
4845:
4842:
4840:
4837:
4835:
4832:
4830:
4829:Chinchilla AI
4827:
4825:
4822:
4820:
4817:
4815:
4812:
4810:
4807:
4805:
4802:
4800:
4797:
4795:
4792:
4790:
4787:
4785:
4782:
4780:
4777:
4775:
4772:
4768:
4765:
4764:
4763:
4760:
4758:
4755:
4753:
4750:
4748:
4745:
4743:
4740:
4738:
4735:
4734:
4732:
4728:
4722:
4719:
4715:
4712:
4710:
4707:
4706:
4705:
4702:
4698:
4695:
4693:
4690:
4688:
4685:
4684:
4683:
4680:
4678:
4675:
4673:
4670:
4668:
4665:
4663:
4660:
4658:
4655:
4653:
4650:
4648:
4645:
4643:
4640:
4638:
4635:
4634:
4632:
4628:
4625:
4621:
4615:
4612:
4610:
4607:
4605:
4602:
4600:
4597:
4595:
4592:
4590:
4587:
4585:
4582:
4581:
4579:
4575:
4569:
4566:
4564:
4561:
4559:
4556:
4554:
4551:
4549:
4546:
4545:
4543:
4539:
4531:
4528:
4527:
4526:
4523:
4521:
4518:
4516:
4513:
4509:
4508:Deep learning
4506:
4505:
4504:
4501:
4497:
4494:
4493:
4492:
4489:
4488:
4486:
4482:
4476:
4473:
4471:
4468:
4464:
4461:
4460:
4459:
4456:
4454:
4451:
4447:
4444:
4442:
4439:
4437:
4434:
4433:
4432:
4429:
4427:
4424:
4422:
4419:
4417:
4414:
4412:
4409:
4407:
4404:
4402:
4399:
4397:
4396:Hallucination
4394:
4390:
4387:
4386:
4385:
4382:
4380:
4377:
4373:
4370:
4369:
4368:
4365:
4364:
4362:
4358:
4352:
4349:
4347:
4344:
4342:
4339:
4337:
4334:
4332:
4329:
4327:
4324:
4322:
4319:
4317:
4314:
4312:
4311:
4307:
4306:
4304:
4302:
4298:
4289:
4284:
4282:
4277:
4275:
4270:
4269:
4266:
4254:
4246:
4244:
4236:
4235:
4232:
4226:
4223:
4221:
4218:
4216:
4213:
4211:
4208:
4206:
4203:
4200:
4196:
4195:
4193:
4189:
4178:
4175:
4174:
4172:
4168:
4161:
4158:
4155:
4152:
4149:
4146:
4143:
4140:
4137:
4134:
4131:
4128:
4127:
4125:
4121:
4114:
4111:
4108:
4105:
4102:
4099:
4098:
4096:
4092:
4089:
4087:Generative AI
4085:
4075:
4072:
4070:
4067:
4065:
4062:
4061:
4059:
4055:
4048:
4045:
4042:
4039:
4036:
4033:
4032:
4030:
4026:
4023:
4019:
4008:
4007:AlphaGeometry
4005:
4002:
3999:
3996:
3993:
3990:
3987:
3986:
3984:
3980:
3969:
3968:
3964:
3961:
3960:
3956:
3955:
3953:
3949:
3942:
3939:
3936:
3933:
3930:
3927:
3926:
3924:
3920:
3913:
3910:
3907:
3904:
3901:
3898:
3895:
3892:
3889:
3886:
3885:
3883:
3879:
3876:
3872:
3869:
3865:
3859:
3856:
3854:
3851:
3849:
3846:
3845:
3842:
3838:
3831:
3826:
3824:
3819:
3817:
3812:
3811:
3808:
3801:
3795:
3791:
3786:
3781:
3776:
3771:
3767:
3763:
3759:
3753:
3750:
3747:
3744:
3741:
3738:
3735:
3731:
3728:
3726:
3722:
3719:
3717:
3714:
3713:
3709:
3704:
3700:
3697:
3693:
3690:
3686:
3685:
3681:
3666:
3662:
3656:
3653:
3641:
3637:
3631:
3629:
3625:
3614:
3610:
3606:
3602:
3596:
3593:
3588:
3584:
3579:
3574:
3569:
3564:
3560:
3556:
3552:
3545:
3542:
3537:
3533:
3528:
3523:
3518:
3513:
3509:
3505:
3501:
3494:
3491:
3486:
3482:
3477:
3472:
3468:
3464:
3460:
3456:
3452:
3448:
3444:
3437:
3434:
3429:
3425:
3421:
3417:
3413:
3409:
3404:
3399:
3395:
3391:
3384:
3377:
3374:
3363:
3359:
3355:
3351:
3345:
3342:
3337:
3333:
3329:
3325:
3321:
3317:
3313:
3309:
3304:
3299:
3295:
3291:
3287:
3280:
3277:
3273:
3269:
3265:
3260:
3257:
3252:
3248:
3242:
3240:
3236:
3231:
3227:
3223:
3219:
3214:
3209:
3205:
3201:
3197:
3193:
3189:
3182:
3179:
3174:
3170:
3164:
3161:
3150:
3146:
3139:
3136:
3131:
3127:
3121:
3118:
3106:
3102:
3096:
3093:
3082:
3078:
3072:
3069:
3066:
3061:
3058:
3055:
3050:
3047:
3043:
3040:(tweet), via
3039:
3035:
3030:
3027:
3016:
3012:
3005:
3002:
2990:
2986:
2979:
2976:
2965:
2961:
2957:
2953:
2946:
2943:
2932:
2928:
2925:Knapp, Alex.
2921:
2918:
2914:
2910:
2904:
2901:
2897:
2893:
2889:
2884:
2881:
2877:
2876:
2871:
2865:
2862:
2858:
2857:
2852:
2846:
2843:
2839:
2838:
2833:
2829:
2824:
2821:
2817:
2813:
2807:
2806:
2801:
2798:Tom Whipple,
2795:
2792:
2788:
2787:
2786:New Scientist
2782:
2777:
2774:
2770:
2769:
2768:New Scientist
2764:
2758:
2755:
2751:
2747:
2741:
2738:
2734:
2730:
2725:
2722:
2717:
2713:
2709:
2705:
2700:
2695:
2691:
2687:
2683:
2676:
2673:
2667:
2664:
2657:
2654:
2642:
2638:
2632:
2629:
2616:
2612:
2606:
2604:
2602:
2598:
2587:
2583:
2577:
2574:
2562:
2561:Bloomberg.com
2558:
2552:
2550:
2546:
2541:
2537:
2532:
2527:
2523:
2519:
2515:
2511:
2506:
2501:
2497:
2493:
2489:
2482:
2479:
2476:
2472:
2468:
2464:
2460:
2456:
2452:
2446:
2443:
2439:
2433:
2430:
2425:
2421:
2417:
2413:
2409:
2405:
2400:
2395:
2391:
2387:
2383:
2376:
2373:
2361:
2357:
2351:
2348:
2336:
2332:
2325:
2323:
2319:
2315:
2310:
2307:
2304:
2299:
2296:
2292:
2287:
2284:
2278:
2273:
2269:
2265:
2260:
2255:
2251:
2247:
2239:
2236:
2225:
2221:
2217:
2213:
2206:
2203:
2198:
2194:
2190:
2186:
2181:
2176:
2172:
2168:
2164:
2160:
2156:
2149:
2146:
2142:
2138:
2132:
2129:
2125:
2121:
2120:the blog post
2115:
2111:
2107:
2104:
2100:
2096:
2090:
2086:
2080:
2077:
2071:
2068:
2064:
2060:
2054:
2052:
2050:
2048:
2044:
2040:
2039:
2034:
2031:Jeremy Kahn,
2028:
2026:
2022:
2011:
2007:
2003:
2002:The Economist
1999:
1993:
1991:
1987:
1982:
1978:
1973:
1968:
1964:
1960:
1956:
1952:
1948:
1944:
1940:
1936:
1932:
1925:
1923:
1919:
1915:
1911:
1906:
1903:
1900:
1896:
1892:
1889:
1888:
1883:
1877:
1873:
1869:
1865:
1861:
1857:
1853:
1848:
1845:
1833:
1829:
1823:
1821:
1819:
1815:
1803:
1799:
1793:
1791:
1789:
1785:
1774:
1770:
1764:
1761:
1750:
1746:
1740:
1737:
1732:
1728:
1723:
1718:
1713:
1708:
1704:
1700:
1696:
1692:
1688:
1680:
1678:
1674:
1669:
1665:
1658:
1655:
1651:
1645:
1642:
1637:
1633:
1629:
1625:
1621:
1617:
1613:
1609:
1605:
1601:
1594:
1592:
1590:
1588:
1586:
1584:
1580:
1576:
1570:
1568:
1566:
1564:
1562:
1560:
1556:
1552:
1551:
1546:
1540:
1538:
1536:
1534:
1532:
1530:
1528:
1526:
1522:
1510:
1506:
1501:
1496:
1492:
1488:
1481:
1478:
1467:
1464:
1457:
1454:
1443:
1439:
1433:
1431:
1429:
1427:
1425:
1423:
1419:
1407:
1403:
1397:
1394:
1387:
1382:
1379:
1377:
1376:AlphaGeometry
1374:
1372:
1369:
1367:
1364:
1362:
1359:
1357:
1354:
1352:
1349:
1347:
1346:IBM Blue Gene
1344:
1342:
1339:
1338:
1333:
1328:
1324:
1320:
1317:
1316:
1311:
1307:
1305:
1301:
1297:
1294:
1293:
1288:
1284:
1280:
1278:
1274:
1270:
1267:
1263:
1259:
1255:
1254:
1250:
1248:
1246:
1242:
1238:
1237:
1232:
1228:
1227:spike protein
1224:
1220:
1216:
1208:
1202:
1199:
1195:
1191:
1188:
1184:
1181:
1178:
1175:
1171:
1168:
1164:
1163:
1162:
1156:
1154:
1151:
1149:
1144:
1142:
1138:
1134:
1131:
1127:
1123:
1119:
1109:
1105:
1102:
1099:
1097:
1093:
1090:Miscellaneous
1088:
1084:
1082:
1078:
1073:
1069:
1065:
1062:
1059:
1055:
1050:
1047:
1043:
1039:
1037:
1033:
1028:
1024:
1022:
1018:
1014:
1008:
1003:
994:
992:
985:
983:
981:
977:
973:
968:
966:
962:
961:
956:
955:
950:
949:
948:New Scientist
944:
943:
938:
937:
932:
931:
925:
922:
919:
915:
911:
907:
903:
895:
893:
887:
885:
883:
879:
875:
871:
866:
864:
860:
854:
852:
847:
843:
838:
836:
832:
827:
821:
819:
817:
813:
808:
806:
802:
797:
795:
787:
777:
767:
760:
758:
755:
753:
747:
745:
741:
738:and selected
737:
733:
729:
725:
721:
717:
713:
708:
703:
701:
697:
693:
691:
687:
683:
679:
675:
670:
668:
662:
660:
651:
643:
639:
634:
632:
630:
626:
621:
617:
613:
608:
606:
604:
600:
596:
595:deep learning
592:
584:
582:
580:
579:deep learning
576:
572:
568:
564:
559:
557:
553:
549:
544:
542:
538:
534:
530:
523:
518:
513:
509:
501:
499:
497:
493:
489:
485:
481:
476:
474:
470:
466:
465:
460:
456:
450:
448:
443:
439:
434:
432:
431:deep learning
428:
424:
420:
416:
412:
401:
396:
394:
389:
387:
382:
381:
379:
378:
371:
368:
367:
361:
360:
353:
350:
348:
345:
343:
340:
338:
335:
334:
331:
326:
325:
318:
315:
313:
310:
308:
305:
303:
300:
298:
294:
291:
289:
286:
284:
281:
279:
276:
275:
272:
267:
266:
259:
256:
254:
251:
249:
246:
244:
241:
237:
236:Mental health
234:
233:
232:
229:
227:
224:
222:
219:
215:
212:
210:
207:
205:
202:
201:
200:
199:Generative AI
197:
195:
192:
190:
187:
185:
182:
180:
177:
176:
173:
168:
167:
160:
157:
155:
152:
150:
147:
145:
142:
140:
139:Deep learning
137:
135:
132:
130:
127:
126:
120:
119:
112:
109:
107:
104:
102:
99:
97:
94:
92:
89:
87:
84:
82:
79:
77:
74:
72:
69:
67:
64:
63:
60:
55:
54:
48:
44:
43:
40:
36:
32:
31:
19:
5018:Hugging Face
4982:David Silver
4676:
4630:Audio–visual
4484:Applications
4463:Augmentation
4308:
4220:Google Pixel
3988:
3965:
3957:
3922:Competitions
3900:AlphaGo Zero
3853:Google Brain
3765:
3761:
3668:. Retrieved
3664:
3655:
3644:. Retrieved
3639:
3616:. Retrieved
3604:
3595:
3561:(22): 7462.
3558:
3554:
3544:
3507:
3503:
3493:
3450:
3446:
3436:
3393:
3389:
3376:
3365:. Retrieved
3354:Fast Company
3353:
3344:
3293:
3289:
3279:
3259:
3250:
3195:
3191:
3181:
3172:
3163:
3152:. Retrieved
3148:
3138:
3129:
3120:
3108:. Retrieved
3104:
3095:
3084:. Retrieved
3080:
3071:
3060:
3049:
3029:
3018:. Retrieved
3014:
3004:
2993:. Retrieved
2991:(in Spanish)
2988:
2978:
2967:. Retrieved
2956:The Guardian
2955:
2945:
2934:. Retrieved
2930:
2920:
2913:Folding@home
2903:
2883:
2873:
2864:
2856:The Guardian
2854:
2845:
2835:
2823:
2811:
2803:
2794:
2784:
2776:
2766:
2757:
2740:
2724:
2689:
2685:
2675:
2666:
2656:
2644:. Retrieved
2640:
2631:
2619:. Retrieved
2614:
2589:. Retrieved
2585:
2576:
2565:. Retrieved
2563:. 2020-11-30
2560:
2495:
2491:
2481:
2462:
2458:
2450:
2445:
2432:
2389:
2385:
2375:
2363:. Retrieved
2359:
2350:
2338:. Retrieved
2335:The Guardian
2334:
2309:
2298:
2286:
2249:
2238:
2227:. Retrieved
2215:
2205:
2162:
2158:
2148:
2136:
2135:John Jumper
2131:
2102:
2099:equivariance
2091:, slide 12).
2084:
2079:
2070:
2058:
2036:
2013:. Retrieved
2001:
1938:
1934:
1905:
1890:
1885:
1878:(blog post).
1863:
1859:
1854:(May 2019),
1847:
1836:. Retrieved
1831:
1806:. Retrieved
1804:. 2024-05-08
1801:
1776:. Retrieved
1772:
1763:
1752:. Retrieved
1748:
1739:
1694:
1690:
1667:
1657:
1644:
1603:
1599:
1548:
1512:. Retrieved
1490:
1480:
1469:. Retrieved
1465:
1456:
1445:. Retrieved
1441:
1409:. Retrieved
1405:
1396:
1356:Rosetta@home
1341:Folding@home
1322:
1321:John Jumper
1313:
1309:
1308:John Jumper
1295:
1290:
1282:
1268:
1265:
1257:
1234:
1212:
1209:Applications
1196:and co- and
1160:
1152:
1145:
1117:
1115:
1067:Download URL
989:
969:
958:
952:
946:
940:
934:
928:
926:
899:
891:
867:
855:
839:
828:
825:
809:
798:
791:
761:Competitions
756:
748:
711:
709:
707:
698:
694:
671:
663:
658:
656:
638:
628:
615:
614:
612:
588:
560:
545:
539:to form the
527:
522:polypeptides
477:
462:
451:
435:
410:
409:
283:Chinese room
172:Applications
5166:Categories
5114:Autoencoder
5069:Transformer
4937:Alex Graves
4885:OpenAI Five
4789:IBM Watsonx
4411:Convolution
4389:Overfitting
4215:Google Labs
4041:Transformer
3015:The Skeptic
2888:Tim Hubbard
2849:Ian Sample,
2646:30 November
2621:30 November
2365:30 November
2340:30 November
2112:2020; also
1411:30 November
1402:"AlphaFold"
1327:AlphaFold 2
1157:Limitations
986:Source code
972:John Jumper
916:winner and
914:Nobel Prize
870:protein NMR
851:side chains
812:open-source
712:AlphaFold 3
674:transformer
659:AlphaFold 2
625:contact map
616:AlphaFold 1
581:technique.
531:consist of
312:Turing test
288:Friendly AI
59:Major goals
18:AlphaFold 2
5192:Categories
5155:Technology
5008:EleutherAI
4967:Fei-Fei Li
4962:Yann LeCun
4875:Q-learning
4858:Decisional
4784:IBM Watson
4692:Midjourney
4584:TensorFlow
4431:Activation
4384:Regression
4379:Clustering
4142:Chinchilla
4069:TensorFlow
3967:The MANIAC
3670:2020-12-01
3646:2020-12-01
3618:2020-12-01
3367:2023-01-24
3303:2204.06860
3154:2021-07-29
3086:2021-07-29
3020:2024-05-12
2995:2024-05-12
2969:2024-05-09
2936:2024-05-09
2896:medium.com
2591:2020-11-30
2567:2020-11-30
2449:Wei Zheng
2392:(3): 249.
2229:2024-05-09
2065:, slide 10
2015:2020-11-30
1893:, 627–628
1838:2020-11-30
1808:2024-05-09
1778:2021-07-24
1754:2021-07-24
1471:2020-11-30
1447:2020-11-30
1388:References
1215:SARS-CoV-2
1010:Data types
954:prediction
506:See also:
502:Background
490:, various
317:Regulation
271:Philosophy
226:Healthcare
221:Government
123:Approaches
5038:MIT CSAIL
5003:Anthropic
4972:Andrew Ng
4870:AlphaZero
4714:VideoPoet
4677:AlphaFold
4614:MindSpore
4568:SpiNNaker
4563:Memristor
4470:Diffusion
4446:Rectifier
4426:Batchnorm
4406:Attention
4401:Adversary
4160:VideoPoet
4101:Assistant
3995:AlphaStar
3989:AlphaFold
3935:Lee Sedol
3906:AlphaZero
3837:Google AI
3752:ColabFold
3613:1059-1028
3555:Molecules
3467:1367-5931
3428:240228913
3412:1545-9985
3362:1085-9241
3328:0031-9007
3230:251159714
2964:0261-3077
2828:Cade Metz
2812:The Times
2805:The Times
2716:254660427
2522:0887-3585
2424:212403708
2408:1548-7105
2268:1476-4687
2224:0362-4331
2197:248156195
2118:See also
2010:0013-0613
1963:1476-4687
1636:227243204
1509:247206999
1366:AlphaZero
1194:cofactors
1167:complexes
1110:automatic
1101:CC-BY 4.0
1021:Organisms
896:Reception
585:Algorithm
480:complexes
473:proteomes
411:AlphaFold
347:AI winter
248:Military
111:AI safety
5146:Portals
4905:Auto-GPT
4737:Word2vec
4541:Hardware
4458:Datasets
4360:Concepts
4243:Category
4191:See also
4094:Chatbots
4001:AlphaDev
3881:Versions
3794:35637307
3640:Deepmind
3587:38005184
3578:10672856
3536:36424442
3485:19700364
3420:34716446
3336:38072605
3222:35902752
3173:Deepmind
2708:36513827
2615:Deepmind
2540:30985027
2416:32132733
2360:Deepmind
2277:11168924
2189:35418629
1981:34265844
1832:Deepmind
1731:34265844
1628:33257889
1514:25 March
1406:Deepmind
1334:See also
1298:706–710
1219:COVID-19
1176:regions.
1148:InterPro
1133:proteome
1122:EMBL-EBI
1040:EMBL-EBI
1012:captured
974:won the
796:(CASP).
772:predict.
720:Alphabet
620:residues
529:Proteins
457:for the
433:system.
423:Alphabet
419:DeepMind
370:Glossary
364:Glossary
342:Progress
337:Timeline
297:Takeover
258:Projects
231:Industry
194:Finance
184:Deepfake
134:Symbolic
106:Robotics
81:Planning
5028:Meta AI
4865:AlphaGo
4849:PanGu-ÎŁ
4819:ChatGPT
4794:Granite
4742:Seq2seq
4721:Whisper
4642:WaveNet
4637:AlexNet
4609:Flux.jl
4589:PyTorch
4441:Sigmoid
4436:Softmax
4301:General
4253:Commons
4107:Sparrow
4035:WaveNet
3959:AlphaGo
3929:Fan Hui
3888:AlphaGo
3874:AlphaGo
3800:version
3785:9184281
3742:website
3740:CASP 14
3665:Fortune
3527:9911346
3476:2749913
3308:Bibcode
3200:Bibcode
3110:27 July
3042:twitter
2989:El PaĂs
2531:6800999
2500:bioRxiv
2252:: 1–3.
2167:Bibcode
2114:website
2110:NeurIPS
2038:Fortune
1972:8371605
1943:Bibcode
1722:8371605
1699:Bibcode
1608:Bibcode
1550:Science
1371:AlphaGo
1130:UniProt
1096:License
1057:Website
1030:Contact
1005:Content
965:misfold
960:folding
936:Science
740:ligands
492:ligands
445:CASP's
352:AI boom
330:History
253:Physics
5043:Huawei
5023:OpenAI
4925:People
4895:MuZero
4757:Gemini
4752:Claude
4687:DALL-E
4599:Theano
4179:(2024)
4162:(2024)
4156:(2023)
4154:Gemini
4150:(2022)
4144:(2022)
4138:(2021)
4132:(2018)
4115:(2023)
4113:Gemini
4109:(2022)
4103:(2016)
4049:(2022)
4043:(2017)
4037:(2016)
4009:(2024)
4003:(2023)
3997:(2019)
3991:(2018)
3970:(2023)
3962:(2017)
3943:(2017)
3941:Ke Jie
3937:(2016)
3931:(2015)
3914:(2019)
3912:MuZero
3908:(2017)
3902:(2017)
3896:(2016)
3894:Master
3890:(2015)
3848:Google
3792:
3782:
3725:GitHub
3611:
3585:
3575:
3534:
3524:
3483:
3473:
3465:
3426:
3418:
3410:
3360:
3334:
3326:
3228:
3220:
3192:Nature
2962:
2931:Forbes
2816:thread
2714:
2706:
2686:Nature
2586:GitHub
2538:
2528:
2520:
2502:
2475:slides
2473:; and
2422:
2414:
2406:
2274:
2266:
2250:Nature
2222:
2195:
2187:
2159:Nature
2137:et al.
2085:et al.
2059:et al.
2008:
1979:
1969:
1961:
1935:Nature
1914:Foldit
1887:Nature
1802:Google
1749:GitHub
1729:
1719:
1691:Nature
1634:
1626:
1600:Nature
1507:
1351:Foldit
1323:et al.
1310:et al.
1292:Nature
1283:et al.
1258:et al.
1243:using
1204:knots.
1187:native
1126:models
1052:Access
945:, and
930:Nature
888:CASP15
882:domain
822:CASP14
816:GitHub
788:CASP13
535:which
494:, and
464:Nature
413:is an
302:Ethics
5109:Mamba
4880:SARSA
4844:LLaMA
4839:BLOOM
4824:GPT-J
4814:GPT-4
4809:GPT-3
4804:GPT-2
4799:GPT-1
4762:LaMDA
4594:Keras
4170:Other
4136:LaMDA
4057:Other
3982:Other
3605:Wired
3424:S2CID
3386:(PDF)
3298:arXiv
3266:, by
3226:S2CID
2712:S2CID
2451:et al
2420:S2CID
2193:S2CID
2103:et al
2095:SE(3)
1632:S2CID
1505:S2CID
1236:ORF3a
1075:Tools
726:with
667:AMBER
577:(AI)
214:Music
209:Audio
5033:Mila
4834:PaLM
4767:Bard
4747:BERT
4730:Text
4709:Sora
4177:Vids
4148:PaLM
4130:BERT
4047:Gato
3790:PMID
3609:ISSN
3583:PMID
3532:PMID
3481:PMID
3463:ISSN
3416:PMID
3408:ISSN
3358:ISSN
3332:PMID
3324:ISSN
3218:PMID
3112:2021
2960:ISSN
2733:CASP
2704:PMID
2648:2020
2623:2020
2536:PMID
2518:ISSN
2436:See
2412:PMID
2404:ISSN
2367:2020
2342:2020
2264:ISSN
2220:ISSN
2185:PMID
2006:ISSN
1977:PMID
1959:ISSN
1727:PMID
1624:PMID
1516:2022
1466:CNBC
1413:2020
1116:The
902:CASP
744:ions
742:and
682:edge
603:GPUs
593:, a
567:CASP
554:and
510:and
496:ions
4774:NMT
4657:OCR
4652:HWR
4604:JAX
4558:VPU
4553:TPU
4548:IPU
4372:SGD
3798:),
3780:PMC
3770:doi
3732:at
3723:on
3573:PMC
3563:doi
3522:PMC
3512:doi
3471:PMC
3455:doi
3398:doi
3316:doi
3294:131
3208:doi
3196:608
2694:doi
2690:613
2526:PMC
2510:doi
2467:doi
2394:doi
2272:PMC
2254:doi
2175:doi
2163:604
1967:PMC
1951:doi
1939:596
1895:doi
1891:577
1868:doi
1717:PMC
1707:doi
1695:596
1616:doi
1604:588
1495:doi
1300:doi
1296:577
1273:doi
1085:yes
1081:Web
1070:yes
967:).
904:'s
814:on
732:RNA
728:DNA
629:how
571:GDT
488:RNA
484:DNA
204:Art
5194::
3788:.
3778:.
3766:19
3764:.
3760:.
3663:.
3638:.
3627:^
3607:.
3603:.
3581:.
3571:.
3559:28
3557:.
3553:.
3530:.
3520:.
3508:20
3506:.
3502:.
3479:.
3469:.
3461:.
3451:13
3445:.
3422:.
3414:.
3406:.
3394:28
3392:.
3388:.
3356:.
3352:.
3330:.
3322:.
3314:.
3306:.
3292:.
3288:.
3249:.
3238:^
3224:.
3216:.
3206:.
3194:.
3190:.
3171:.
3147:.
3128:.
3103:.
3079:.
3036:,
3013:.
2987:.
2958:.
2954:.
2929:.
2911:,
2894:,
2890:,
2872:.
2853:,
2834:,
2830:,
2818:).
2802:,
2783:,
2765:,
2748:,
2710:.
2702:.
2688:.
2684:.
2661:Ă….
2639:.
2613:.
2600:^
2584:.
2559:.
2548:^
2534:.
2524:.
2516:.
2508:.
2496:87
2494:.
2490:.
2463:87
2461:,
2457:,
2418:.
2410:.
2402:.
2390:17
2388:.
2384:.
2358:.
2333:.
2321:^
2270:.
2262:.
2248:.
2218:.
2214:.
2191:.
2183:.
2173:.
2161:.
2157:.
2108:,
2046:^
2035:,
2024:^
2000:.
1989:^
1975:.
1965:.
1957:.
1949:.
1937:.
1933:.
1921:^
1912:,
1884:,
1864:35
1862:,
1858:,
1830:.
1817:^
1800:.
1787:^
1771:.
1747:.
1725:.
1715:.
1705:.
1693:.
1689:.
1676:^
1666:.
1630:.
1622:.
1614:.
1602:.
1582:^
1558:^
1547:,
1524:^
1503:.
1493:.
1489:.
1440:.
1421:^
1404:.
1289:,
1269:87
1264:,
939:,
933:,
746:.
734:,
730:,
605:.
599:AI
550:,
498:.
486:,
475:.
4287:e
4280:t
4273:v
4201:"
4197:"
3829:e
3822:t
3815:v
3796:.
3772::
3754:(
3673:.
3649:.
3621:.
3589:.
3565::
3538:.
3514::
3487:.
3457::
3430:.
3400::
3370:.
3338:.
3318::
3310::
3300::
3253:.
3232:.
3210::
3202::
3157:.
3132:.
3114:.
3089:.
3023:.
2998:.
2972:.
2939:.
2718:.
2696::
2650:.
2625:.
2594:.
2570:.
2542:.
2512::
2469::
2453:,
2426:.
2396::
2369:.
2344:.
2280:.
2256::
2232:.
2199:.
2177::
2169::
2097:-
2018:.
1983:.
1953::
1945::
1897::
1870::
1841:.
1811:.
1781:.
1757:.
1733:.
1709::
1701::
1670:.
1638:.
1618::
1610::
1518:.
1497::
1474:.
1450:.
1415:.
1302::
1275::
1169:.
846:Ă…
776:Ă…
399:e
392:t
385:v
295:/
20:)
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.
