Protein sequence‐to‐structure learning: Is this the end(‐to‐end revolution)?

Laine, Elodie; Eismann, Stephan; Elofsson, Arne; Grudinin, Sergei

doi:10.1002/prot.26235

Cited by 44Protein Structure and DynamicsMachine Learning in BioinformaticsEnzyme Structure and Function

Protein sequence‐to‐structure learning: Is this the end(‐to‐end revolution)?

Open access

Laine, Elodie, Eismann, Stephan, Elofsson, Arne, Grudinin, Sergei·2021·Proteins: Structure, Function, and Bioinformatics

DOI 10.1002/prot.26235

Open this paper in CoChat — free account Open PDF

Abstract

The potential of deep learning has been recognized in the protein structure prediction community for some time, and became indisputable after CASP13. In CASP14, deep learning has boosted the field to unanticipated levels reaching near‐experimental accuracy. This success comes from advances transferred from other machine learning areas, as well as methods specifically designed to deal with protein sequences and structures, and their abstractions. Novel emerging approaches include (i) geometric learning, that is, learning on representations such as graphs, three‐dimensional (3D) Voronoi tessellations, and point clouds; (ii) pretrained protein language models leveraging attention; (iii) equivariant architectures preserving the symmetry of 3D space; (iv) use of large meta‐genome databases; (v) combinations of protein representations; and (vi) finally truly end‐to‐end architectures, that is, differentiable models starting from a sequence and returning a 3D structure. Here, we provide an overview and our opinion of the novel deep learning approaches developed in the last 2 years and widely used in CASP14.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

Integrity signals

Registry

Crossref · 10.1002/prot.26235

Cites

174 of 174 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

The improvements in model accuracy by AlphaFold2 and the other leading groups almost all arise from more advanced use of deep learning methods.
“The improvements in model accuracy by AlphaFold2 and the other leading groups almost all arise from more advanced use of deep learning methods, discussed in 37.”
From Critical assessment of methods of protein structure prediction (CASP)—Round <scp>XIV</scp> (2021) · cites this paper as [37]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Attention Is All You Need
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 2× · see citing works
“In particular, advances in the treatment of language [25] and of 3D geometry [26, 27, 28, 29, 30] by deep learning architectures have further beneﬁted the ﬁeld of protein structure and function prediction.”
Background
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 48,293 · see citing works
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764 · cited 7× · see citing works
“This observation was exploited by AlphaFold2 through the implementation of triangular self-attention.(31) In the graph on the left, the red arrows indicate standard convolution aggregating information from neighboring nodes.”
BackgroundMethodInfluential
Semi-supervised classification with graph convolutional networks
Kipf, T. N., & Welling, M. (2016) · arXiv · doi:10.48550/arxiv.1609.02907 · Cited by 24,853 · see citing works
Multi-scale context aggregation by dilated convolutions
Yu, F., & Koltun, V. (2015) · arXiv · doi:10.48550/arxiv.1511.07122 · Cited by 7,738 · cited 2× · see citing works
“[110] Yu F, Koltun V, Multi-Scale Context Aggregation by Dilated Convolutions; 2016.”
Background
Sparse inverse covariance estimation with the graphical lasso
Friedman, J., Hastie, T., & Tibshirani, R. (2008) · Biostatistics · doi:10.1093/biostatistics/kxm045 · Cited by 6,355 · cited 1× · see citing works
“Over the past 10 years, direct coupling analysis (DCA)-based models[12], including Potts model and pseudolikelihood maximization [8, 81, 82, 83], and Graphical lasso-based (low-rank) models [84, 85, 86] became widespread in the community.”
Accurate prediction of protein structures and interactions using a 3-track network
Baek, M., DiMaio, F., Anishchenko, I., Dauparas, J., Ovchinnikov, S., Lee, G. R., Wang, J., Cong, Q., Kinch, L. N., Schaeffer, R. D., et al. (2021) · doi:10.1101/2021.06.14.448402 · Cited by 5,542 · cited 1× · see citing works
“combination was also leveraged in the most recent RoseTTAFold network.(148)”
Background
Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces
Behler, J., & Parrinello, M. (2007) · Physical Review Letters · doi:10.1103/physrevlett.98.146401 · Cited by 5,010 · cited 1× · see citing works
“These architectures can be seen as equivariant extensions of neural-network-based radial and angular symmetry functions for molecular structure [124, 125, 126].”
Background
A fast algorithm for particle simulations
Greengard, L., & Rokhlin, V. (1987) · Journal of Computational Physics · doi:10.1016/0021-9991(87)90140-9 · Cited by 4,967 · cited 1× · see citing works
“Spherical harmonics have played a prominent role in molecular surface representations for several decades(144,145) and are also at the heart of the classical fast multipole method.(146)”
Background
Improved protein structure prediction using potentials from deep learning
Senior, A. W., Evans, R., Jumper, J., Kirkpatrick, J., Sifre, L., Green, T., Qin, C., Žídek, A., Nelson, A. W. R., Bridgland, A., et al. (2020) · Nature · doi:10.1038/s41586-019-1923-7 · Cited by 3,378 · cited 3× · see citing works
“Stacking dilated convolutions with increasingly large d allows operating on exponentially large receptive fields, while retaining short backpropagations [126, 127, 18].”
Background
Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences
Rives, A., Meier, J., Sercu, T., Goyal, S., Lin, Z., Liu, J., Guo, D., Ott, M., Zitnick, C. L., Ma, J., et al. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2016239118 · Cited by 3,286 · see citing works
Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences
Rives, A., Meier, J., Sercu, T., Goyal, S., Lin, Z., Liu, J., Guo, D., Ott, M., Zitnick, C. L., Ma, J., et al. (2019) · doi:10.1101/622803 · Cited by 3,280 · cited 1× · see citing works
“Along this line, recent works have shown that NLP models pre-trained on millions of unlabelled protein sequences can be eﬀectively ﬁne-tuned with small amount of labelled data toward accurately predicting mutational eﬀects as well as 3D contacts [185, 184, 186].”
Background
The Energy Landscapes and Motions of Proteins
Frauenfelder, H., Sligar, S. G., & Wolynes, P. G. (1991) · Science · doi:10.1126/science.1749933 · Cited by 3,250 · cited 1× · see citing works
“Flexibility is necessary for binding, enzymatic reactions, transport, and many more.(149,150) Many proteins adopt two or”
Background
On representing chemical environments
Bartók, A. P., Kondor, R., & Csányi, G. (2013) · Physical Review B · doi:10.1103/physrevb.87.184115 · Cited by 2,572 · cited 1× · see citing works
“In particular, advances in the treatment of language [25] and of 3D geometry [26, 27, 28, 29, 30] by deep learning architectures have further benefited the field of protein structure and function prediction.”
Background
Dynamic personalities of proteins
Henzler-Wildman, K., & Kern, D. (2007) · Nature · doi:10.1038/nature06522 · Cited by 2,545 · cited 1× · see citing works
“Flexibility is necessary for binding, enzymatic reactions, transport, and many more.(149,150) Many proteins adopt two or”
Background
ANI-1: an extensible neural network potential with DFT accuracy at force field computational cost
Smith, J. S., Isayev, O., & Roitberg, A. E. (2017) · Chemical Science · doi:10.1039/c6sc05720a · Cited by 1,921 · cited 1× · see citing works
“These architectures can be seen as equivariant extensions of neural-network-based radial and angular symmetry functions for molecular structure [124, 125, 126].”
Background
Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs
Monti, F., Boscaini, D., Masci, J., Rodola, E., Svoboda, J., & Bronstein, M. M. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.576 · Cited by 1,820 · cited 2× · see citing works
“[106] Monti F, Boscaini D, Masci J, Rodola E, Svoboda J, Bronstein MM.”
Background
Estimation of effective interresidue contact energies from protein crystal structures: quasi-chemical approximation
Miyazawa, S., & Jernigan, R. L. (1985) · Macromolecules · doi:10.1021/ma00145a039 · Cited by 1,588 · see citing works
Direct-coupling analysis of residue coevolution captures native contacts across many protein families
Morcos, F., Pagnani, A., Lunt, B., Bertolino, A., Marks, D. S., Sander, C., Zecchina, R., Onuchic, J. N., Hwa, T., & Weigt, M. (2011) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1111471108 · Cited by 1,434 · cited 1× · see citing works
“They enabled the modelling of 3D structures for large protein families [9, 10, 11, 12, 13, 14].”
Background
Improved protein structure prediction using predicted interresidue orientations
Yang, J., Anishchenko, I., Park, H., Peng, Z., Ovchinnikov, S., & Baker, D. (2020) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1914677117 · Cited by 1,421 · cited 1× · see citing works
“First, as mentioned above, efficient methods to use 2D information for generating 3D models exist.(51,58) Further, the most popular residue- or even atom-level loss functions used in deep neural networks (DNNs) do not depend on the superposition of the predicted”
Method
Big Data: Astronomical or Genomical?
Stephens, Z. D., Lee, S. Y., Faghri, F., Campbell, R. H., Zhai, C., Efron, M. J., Iyer, R., Schatz, M. C., Sinha, S., & Robinson, G. E. (2015) · PLOS Biology · doi:10.1371/journal.pbio.1002195 · Cited by 1,340 · cited 1× · see citing works
“Genomics has become pre-eminent in terms of data scale, with an exponential growth [64, 65].”
Background
Protein 3D Structure Computed from Evolutionary Sequence Variation
Marks, D. S., Colwell, L. J., Sheridan, R., Hopf, T. A., Pagnani, A., Zecchina, R., & Sander, C. (2011) · PLoS ONE · doi:10.1371/journal.pone.0028766 · Cited by 1,144 · cited 2× · see citing works
“They enabled the modelling of 3D structures for large protein families [9, 10, 11, 12, 13, 14].”
BackgroundMethod
lDDT: a local superposition-free score for comparing protein structures and models using distance difference tests
Mariani, V., Biasini, M., Barbato, A., & Schwede, T. (2013) · Bioinformatics · doi:10.1093/bioinformatics/btt473 · Cited by 1,138 · cited 2× · see citing works
“[57] Mariani V, Biasini M, Barbato A, Schwede T.”
BackgroundMethod
Calculation of conformational ensembles from potentials of mena force
Sippl, M. J. (1990) · Journal of Molecular Biology · doi:10.1016/s0022-2836(05)80269-4 · Cited by 1,064 · cited 1× · see citing works
“Historically, machine-learning-based scoring functions [119, 120] were inspired by statistical potentials [121, 122] relying on pairwise distance/ angular distributions or contact maps, which are perfectly rotation”
Background
Identification of direct residue contacts in protein–protein interaction by message passing
Weigt, M., White, R. A., Szurmant, H., Hoch, J. A., & Hwa, T. (2009) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0805923106 · Cited by 1,025 · cited 2× · see citing works
“[7] Weigt M, White RA, Szurmant H, Hoch JA, Hwa T.”
Background
Accurate De Novo Prediction of Protein Contact Map by Ultra-Deep Learning Model
Wang, S., Sun, S., Li, Z., Zhang, R., & Xu, J. (2017) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1005324 · Cited by 931 · cited 2× · see citing works
“Shifting from unsupervised statistical inference to supervised deep learning further boosted the accuracy of the predicted contacts, and extended the applicability of this conceptual framework to families with fewer sequences [15, 16] and to the prediction of residue-residue distances [17, 18].”
Background
Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and GaugesPreprint
Bronstein, M. M., Bruna, J., Cohen, T., & Veličković, P. (2021) · arXiv (Cornell University) · doi:10.4230/oasics.dx.2026.1 · Cited by 928 · see citing works
Clustering huge protein sequence sets in linear time
Steinegger, M., & Söding, J. (2018) · Nature Communications · doi:10.1038/s41467-018-04964-5 · Cited by 875 · cited 1× · see citing works
“They include databases like NCBI GenBank [66], Metaclust [67], BFD [68], MetaEuk [69], EBI MGnify [70], and IMG/M [71].”
Method
Correlated mutations and residue contacts in proteins
Göbel, U., Sander, C., Schneider, R., & Valencia, A. (1994) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.340180402 · Cited by 863 · see citing works
IMG/M v.5.0: an integrated data management and comparative analysis system for microbial genomes and microbiomes
Chen, I.-M. A., Chu, K., Palaniappan, K., Pillay, M., Ratner, A., Huang, J., Huntemann, M., Varghese, N. J., White, J. R., Seshadri, R., et al. (2018) · Nucleic Acids Research · doi:10.1093/nar/gky901 · Cited by 811 · cited 1× · see citing works
“They include databases like NCBI GenBank,(66) Metaclust,(67) BFD,(68) MetaEuk,(69) EBI MGnify,(70) and IMG/M.(71) Since CASP12, several teams attempted to exploit this type of data, mostly to increase the depth of the MSAs and obtain a more accurate estimation of (co-)evolutionary features.”
Background
Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning
Gaínza, P., Sverrisson, F., Monti, F., Rodolà, E., Boscaini, D., Bronstein, M. M., & Correia, B. E. (2019) · Nature Methods · doi:10.1038/s41592-019-0666-6 · Cited by 811 · cited 1× · see citing works
“Other developments include the application of ideas from the recent MoNet manifold network architecture(123) toward learning protein surfaces with very exciting outcomes for protein binding sites and protein–protein complexes prediction.(124,125) Overall, 2D surface manifolds seem very”
Background
Disease variant prediction with deep generative models of evolutionary data
Frazer, J., Notin, P., Dias, M., Gomez, A., Min, J. K., Brock, K., Gal, Y., & Marks, D. S. (2021) · Nature · doi:10.1038/s41586-021-04043-8 · Cited by 791 · see citing works
Mutation effects predicted from sequence co-variation
Hopf, T. A., Ingraham, J. B., Poelwijk, F. J., Schärfe, C. P. I., Springer, M., Sander, C., & Marks, D. S. (2017) · Nature Biotechnology · doi:10.1038/nbt.3769 · Cited by 775 · cited 3× · see citing works
“The success of these methods lies in their ability to capture dependencies between protein residues either by explicitly estimating inter-residue (pairwise) couplings [178, 179] or by implicitly accounting for global sequence contexts [174, 176].”
BackgroundMethod
PSICOV: precise structural contact prediction using sparse inverse covariance estimation on large multiple sequence alignments
Jones, D. T., Buchan, D. W. A., Cozzetto, D., & Pontil, M. (2011) · Bioinformatics · doi:10.1093/bioinformatics/btr638 · Cited by 760 · see citing works
Deviations from Standard Atomic Volumes as a Quality Measure for Protein Crystal Structures
Pontius, J., Richelle, J., & Wodak, S. J. (1996) · Journal of Molecular Biology · doi:10.1006/jmbi.1996.0628 · Cited by 754 · see citing works
GenBank
Benson, D. A., Cavanaugh, M. L., Clark, K., Karsch‐Mizrachi, I., Ostell, J., Pruitt, K. D., & Sayers, E. W. (2017) · Nucleic Acids Research · doi:10.1093/nar/gkx1094 · Cited by 752 · see citing works
Boltzmann generators: Sampling equilibrium states of many-body systems with deep learning
Noé, F., Olsson, S., Köhler, J., & Wu, H. (2019) · Science · doi:10.1126/science.aaw1147 · Cited by 747 · cited 1× · see citing works
“Alternatively, instead of learning from MD trajectories, deep learning can be used to generate conformational ensembles obeying the Boltzmann distribution.(163)”
Method
Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudsPreprint
Thomas, N., Smidt, T., Kearnes, S. M., Yang, L., Li, L., Kohlhoff, K., & Riley, P. F. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1802.08219 · Cited by 726 · cited 5× · see citing works
“In particular, advances in the treatment of language [14] and of 3D geometry [15, 16, 17, 18] by deep learning architectures have further benefited the field of protein structure and function prediction.”
BackgroundInfluential
Evaluating Protein Transfer Learning with TAPE
Rao, R., Bhattacharya, N., Thomas, N., Duan, Y., Chen, X., Canny, J., Abbeel, P., & Song, Y. S. (2019) · doi:10.1101/676825 · Cited by 719 · cited 1× · see citing works
“” During the semi-supervised training phase, the model attempts to predict a masked or the next token.(98) In CASP14, EMBER directly made use of ELMo and BERT while HMS-Casper(53)”
Method
Can correct protein models be identified?
Wallner, B., & Elofsson, A. (2003) · Protein Science · doi:10.1110/ps.0236803 · Cited by 709 · see citing works
Self-citation
Rethinking Attention with Performers
Choromanski, K., Likhosherstov, V., Dohan, D., Song, X., Gane, A., Sarlos, T., Hawkins, P., Davis, J., Mohiuddin, A., Kaiser, L., et al. (2020) · arXiv · doi:10.48550/arxiv.2009.14794 · Cited by 695 · cited 2× · see citing works
“Attention mechanisms have made a major breakthrough in NLP, as they allow keeping in memory the sequence context (although limited in practical applications) in translation tasks [112, 113, 14, 114].”
Background
Assessing the utility of coevolution-based residue–residue contact predictions in a sequence- and structure-rich era
Kamisetty, H., Ovchinnikov, S., & Baker, D. A. (2013) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1314045110 · Cited by 672 · cited 1× · see citing works
“Over the past 10 years, direct coupling analysis (DCA)-based models[12], including Potts model and pseudolikelihood maximization [8, 81, 82, 83], and Graphical lasso-based (low-rank) models [84, 85, 86] became widespread in the community.”
Method
Robust and accurate prediction of residue–residue interactions across protein interfaces using evolutionary information
Ovchinnikov, S., Kamisetty, H., & Baker, D. (2014) · eLife · doi:10.7554/elife.02030 · Cited by 629 · cited 2× · see citing works
“Such strategy has been employed to predict bacterial complexes [144, 145, 146], to single out pairs of interacting paralogs [147, 12] and to gain insight into the interactomes of viruses, like HCV [148], or bacteria, like E.”
De novo protein design by deep network hallucination
Anishchenko, I., Chidyausiku, T. M., Ovchinnikov, S., Pellock, S. J., & Baker, D. (2020) · doi:10.1101/2020.07.22.211482 · Cited by 629 · cited 1× · see citing works
“trained DNNs designed to predict inter-residue distances can be repurposed to estimate the impact of mutations on the 3D structure toward guiding the generation of new sequences predicted to fold to new structures.(186)”
Background
Protein structure prediction from sequence variation
Marks, D. S., Hopf, T. A., & Sander, C. (2012) · Nature Biotechnology · doi:10.1038/nbt.2419 · Cited by 622 · cited 1× · see citing works
“They enabled the modelling of 3D structures for large protein families [9, 10, 11, 12, 13, 14].”
Background
CryoDRGN: reconstruction of heterogeneous cryo-EM structures using neural networks
Zhong, E. D., Bepler, T., Berger, B., & Davis, J. H. (2021) · Nature Methods · doi:10.1038/s41592-020-01049-4 · Cited by 613 · see citing works
Deep generative models of genetic variation capture the effects of mutations
Riesselman, A. J., Ingraham, J. B., & Marks, D. S. (2018) · Nature Methods · doi:10.1038/s41592-018-0138-4 · Cited by 601 · cited 2× · see citing works
“diction and protein design can be unified in a common theoretical framework extracting information from protein sequences.(171,172,175)”
Background
MGnify: the microbiome analysis resource in 2020
Mitchell, A. L., Almeida, A., Beracochea, M., Boland, M., Burgin, J., Cochrane, G., Crusoe, M. R., Kale, V., Potter, S. C., Richardson, L. J., et al. (2019) · Nucleic Acids Research · doi:10.1093/nar/gkz1035 · Cited by 562 · cited 1× · see citing works
“They include databases like NCBI GenBank [66], Metaclust [67], BFD [68], MetaEuk [69], EBI MGnify [70], and IMG/M [71].”
Method
Protein structure determination using metagenome sequence data
Ovchinnikov, S., Park, H., Varghese, N., Huang, P.-S., Pavlopoulos, G. A., Kim, D. E., Kamisetty, H., Kyrpides, N. C., & Baker, D. (2017) · Science · doi:10.1126/science.aah4043 · Cited by 520 · cited 1× · see citing works
“Both methods exploit predicted or observed inter-residue distances to improve the sequence alignments, a strategy that proved powerful already in CASP13.(72,79,80)”
Method
Sequence co-evolution gives 3D contacts and structures of protein complexes
Hopf, T. A., Schärfe, C. P. I., Rodrigues, J. P. G. L. M., Green, A. G., Kohlbacher, O., Sander, C., Bonvin, A. M. J. J., & Marks, D. S. (2014) · eLife · doi:10.7554/elife.03430 · Cited by 515 · cited 2× · see citing works
“Such strategy has been employed to predict bacterial complexes [144, 145, 146], to single out pairs of interacting paralogs [147, 12] and to gain insight into the interactomes of viruses, like HCV [148], or bacteria, like E.”
Background
A Convolutional Encoder Model for Neural Machine Translation
Gehring, J., Auli, M., Grangier, D., & Dauphin, Y. (2016) · Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/p17-1012 · Cited by 470 · see citing works
Protein-level assembly increases protein sequence recovery from metagenomic samples manyfold
Steinegger, M., Mirdita, M., & Söding, J. (2019) · Nature Methods · doi:10.1038/s41592-019-0437-4 · Cited by 445 · cited 1× · see citing works
“They include databases like NCBI GenBank [66], Metaclust [67], BFD [68], MetaEuk [69], EBI MGnify [70], and IMG/M [71].”
Method
Low-N protein engineering with data-efficient deep learning
Biswas, S., Khimulya, G., Alley, E. C., Esvelt, K. M., & Church, G. M. (2021) · Nature Methods · doi:10.1038/s41592-021-01100-y · Cited by 422 · cited 1× · see citing works
“They also hold great potential for guiding protein design and engineering.(182,183) The success of these methods lies in their ability to capture dependencies between protein residues either by explicitly estimating inter-residue (pairwise) couplings(177,178) or by implicitly accounting for global sequence contexts.”
Background
CCMpred—fast and precise prediction of protein residue–residue contacts from correlated mutations
Seemayer, S., Gruber, M., & Söding, J. (2014) · Bioinformatics · doi:10.1093/bioinformatics/btu500 · Cited by 416 · cited 1× · see citing works
“Over the past 10 years, direct coupling analysis (DCA)-based models[12], including Potts model and pseudolikelihood maximization [8, 81, 82, 83], and Graphical lasso-based (low-rank) models [84, 85, 86] became widespread in the community.”
Method
Distance-based protein folding powered by deep learning
Xu, J. (2019) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1821309116 · Cited by 401 · cited 1× · see citing works
“and extended the applicability of this conceptual framework to families with fewer sequences(15,16) and to the prediction of residueresidue distances.(17,18) These advances have significantly increased the protein structure modeling coverage of genomes, and also of bacterial interactomes.”
Background
End-to-End Differentiable Learning of Protein Structure
AlQuraishi, M. (2019) · Cell Systems · doi:10.1016/j.cels.2019.03.006 · Cited by 371 · see citing works
MSA Transformer
Rao, R., Liu, J., Verkuil, R., Meier, J., Canny, J. F., Abbeel, P., Sercu, T., & Rives, A. (2021) · doi:10.1101/2021.02.12.430858 · Cited by 350 · cited 3× · see citing works
“Recent works have also shown that the residue-residue dependencies extracted by certain attention heads in transformers trained on large amounts of sequences can be directly interpreted as 3D contacts or distances [133, 134, 62].”
BackgroundMethod
MetaPSICOV: combining coevolution methods for accurate prediction of contacts and long range hydrogen bonding in proteins
Jones, D. T., Singh, T., Kościółek, T., & Tetchner, S. (2014) · Bioinformatics · doi:10.1093/bioinformatics/btu791 · Cited by 345 · see citing works
Transformer protein language models are unsupervised structure learners
Rao, R., Meier, J., Sercu, T., Ovchinnikov, S., & Rives, A. (2020) · doi:10.1101/2020.12.15.422761 · Cited by 341 · cited 1× · see citing works
“works have also shown that the residue-residue dependencies extracted by certain attention heads in transformers trained on large amounts of sequences can be directly interpreted as 3D contacts or distances.(48,133,134)”
Background
Learning generative models for protein fold families
Balakrishnan, S., Kamisetty, H., Carbonell, J. G., Lee, S., & Langmead, C. J. (2011) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.22934 · Cited by 316 · cited 2× · see citing works
“Over the past 10 years, direct coupling analysis (DCA)-based models[12], including Potts model and pseudolikelihood maximization [8, 81, 82, 83], and Graphical lasso-based (low-rank) models [84, 85, 86] became widespread in the community.”
BackgroundMethod
An evolution-based model for designing chorismate mutase enzymes
Russ, W. P., Figliuzzi, M., Stocker, C., Barrat-Charlaix, P., Socolich, M., Kast, P., Hilvert, D., Monasson, R., Cocco, S., Weigt, M., et al. (2020) · Science · doi:10.1126/science.aba3304 · Cited by 313 · cited 1× · see citing works
“They also hold great potential for guiding protein design and engineering.(182,183) The success of these methods lies in their ability to capture dependencies between protein residues either by explicitly estimating inter-residue (pairwise) couplings(177,178) or by implicitly accounting for global sequence contexts.”
Background
MetaEuk—sensitive, high-throughput gene discovery, and annotation for large-scale eukaryotic metagenomics
Levy Karin, E., Mirdita, M., & Söding, J. (2020) · Microbiome · doi:10.1186/s40168-020-00808-x · Cited by 299 · cited 1× · see citing works
“They include databases like NCBI GenBank [66], Metaclust [67], BFD [68], MetaEuk [69], EBI MGnify [70], and IMG/M [71].”
Method
ProGen: Language Modeling for Protein GenerationPreprint
Madani, A., McCann, B., Naik, N., Keskar, N. S., Anand, N., Eguchi, R. R., Huang, P., & Socher, R. (2020) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2020.03.07.982272 · Cited by 282 · cited 2× · see citing works
“[166] Madani A, McCann B, Naik N, Keskar NS, Anand N, Eguchi RR, et al.”
Background
Coevolutionary Landscape Inference and the Context-Dependence of Mutations in Beta-Lactamase TEM-1
Figliuzzi, M., Jacquier, H., Schug, A. H., Tenaillon, O., & Weigt, M. (2015) · Molecular Biology and Evolution · doi:10.1093/molbev/msv211 · Cited by 263 · cited 2× · see citing works
“The success of these methods lies in their ability to capture dependencies between protein residues either by explicitly estimating inter-residue (pairwise) couplings [178, 179] or by implicitly accounting for global sequence contexts [174, 176].”
BackgroundMethod
Improved protein structure prediction by deep learning irrespective of co-evolution information
(2021) · doi:10.1038/s42256-021-00348-5 · Cited by 256 · cited 2× · see citing works
“For example, RaptorX [36], methods from the Yang and Baker teams [72, 73], Multicom [45], and GALAXY exploited metagenome data for contact prediction and distance estimation between residue pairs in combination with residual convolutional neural networks (resC-NNs).”
Method
Improved protein structure refinement guided by deep learning based accuracy estimation
Hiranuma, N., Park, H., Baek, M., Anishchenko, I., Dauparas, J., & Baker, D. (2021) · Nature Communications · doi:10.1038/s41467-021-21511-x · Cited by 243 · see citing works
Protein interaction networks revealed by proteome coevolution
Cong, Q., Anishchenko, I., Ovchinnikov, S., & Baker, D. (2019) · Science · doi:10.1126/science.aaw6718 · Cited by 241 · cited 1× · see citing works
“Such strategy has been employed to predict bacterial complexes,(22,164,165) to single out pairs of interacting paralogs(23,166) and to gain insight into the interactomes of viruses, like HCV,(167) or bacteria, like Escherichia coli.(24)”
Method
Translating HIV Sequences into Quantitative Fitness Landscapes Predicts Viral Vulnerabilities for Rational Immunogen Design
Ferguson, A. L., Mann, J. K., Omarjee, S., Ndung’u, T., Walker, B. D., & Chakraborty, A. K. (2013) · Immunity · doi:10.1016/j.immuni.2012.11.022 · Cited by 232 · see citing works
3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric DataPreprint
Weiler, M., Geiger, M., Welling, M., Boomsma, W., & Cohen, T. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1807.02547 · Cited by 231 · cited 4× · see citing works
“Volumetric representations 3DCNN[84] A non-invariant 3D CNN Ornate[87] A local frame-based 3D CNN model with learned atom embeddings Sato-3DCNN[88] A local frame-based 3D CNN 3DCNN_prof (P3CMQA)[89] Extends Sato-3DCNN with predicted features and PSSMs SE(3)-3DCNN[85] An invariant 3D CNN based on [86] trained for prote…”
BackgroundMethodInfluential
Genomics-aided structure prediction
Sulkowska, J. I., Morcos, F., Weigt, M., Terence Hwa, B., & Onuchic, J. N. (2012) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1207864109 · Cited by 226 · cited 2× · see citing works
“They enabled the modelling of 3D structures for large protein families [9, 10, 11, 12, 13, 14].”
Background
Large-scale determination of previously unsolved protein structures using evolutionary information
Ovchinnikov, S., Kinch, L., Park, H., Liao, Y., Pei, J., Kim, D. E., Kamisetty, H., Grishin, N. V., & Baker, D. (2015) · eLife · doi:10.7554/elife.09248 · Cited by 222 · cited 2× · see citing works
“These advances have significantly increased the protein structure modelling coverage of genomes [8, 9, 10], and also of bacterial interactomes [11, 12, 13].”
Background
GEMME: A Simple and Fast Global Epistatic Model Predicting Mutational Effects
Laine, É., Karami, Y., & Carbone, A. (2019) · Molecular Biology and Evolution · doi:10.1093/molbev/msz179 · Cited by 197 · cited 1× · see citing works
“The success of these methods lies in their ability to capture dependencies between protein residues either by explicitly estimating inter-residue (pairwise) couplings(177,178) or by implicitly accounting for global sequence contexts.(173,175) In essence, the concepts at play are no different from those implemented for…”
Background
Voronoi and Voronoi-related tessellations in studies of protein structure and interaction.
Poupon, A. (2004) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2004.03.010 · Cited by 192 · cited 1× · see citing works
“The Voronoi tessellation turned out to be a very powerful description of protein structure and interactions and has been used in structural bioinformatics for several decades [112, 113, 114, 115, 116].”
GPCRmd uncovers the dynamics of the 3D-GPCRome
Rodríguez‐Espigares, I., Torrens‐Fontanals, M., Tiemann, J. K. S., Aranda-García, D., Ramírez‐Anguita, J. M., Stępniewski, T. M., Worp, N., Varela‐Rial, A., Morales-Pastor, A., Medel-Lacruz, B., et al. (2020) · Nature Methods · doi:10.1038/s41592-020-0884-y · Cited by 187 · see citing works
Fast pseudolikelihood maximization for direct-coupling analysis of protein structure from many homologous amino-acid sequences
Ekeberg, M., Hartonen, T., & Aurell, E. M. (2014) · Journal of Computational Physics · doi:10.1016/j.jcp.2014.07.024 · Cited by 184 · see citing works
High precision in protein contact prediction using fully convolutional neural networks and minimal sequence features
Jones, D. T., & Kandathil, S. M. (2018) · Bioinformatics · doi:10.1093/bioinformatics/bty341 · Cited by 184 · cited 1× · see citing works
“…to a "Recurrent Geometrical Network" comprising LSTM and geometric units and outputting a 3D structure a variety of of deep-learning models [91, 16, 92, 93, 21, 38, 73, 42, 41, 37, 45], including generative adversarial networks for contact map generation and reﬁne-ment [94, 35], got widely used to…”
Background
DeepMSA: constructing deep multiple sequence alignment to improve contact prediction and fold-recognition for distant-homology proteins
Zhang, C., Zheng, W. H., Mortuza, S. M., Li, Y., & Zhang, Y. (2019) · Bioinformatics · doi:10.1093/bioinformatics/btz863 · Cited by 182 · cited 1× · see citing works
“To gather large amounts of sequences, coming from different sources, many teams relied on the DeepMSA algorithm.(75) Most of the time, the sequences were integrated altogether in a single MSA.”
Method
VoroMQA: Assessment of protein structure quality using interatomic contact areas
Olechnovič, K., & Venclovas, Č. (2017) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.25278 · Cited by 181 · see citing works
Deep learning extends de novo protein modelling coverage of genomes using iteratively predicted structural constraints
Greener, J. G., Kandathil, S. M., & Jones, D. T. (2019) · Nature Communications · doi:10.1038/s41467-019-11994-0 · Cited by 178 · cited 1× · see citing works
“For example, QDeep QA predictions(60) are based on distance estimations from DMPfold.(21) In GalaxyRefine2,(61) RefineD,(62) and Baker suite,(63) the QA is incorporated into a model refinement pipeline.”
Background
Patterns of divergence in homologous proteins as indicators of secondary and tertiary structure: A prediction of the structure of the catalytic domain of protein kinases
Benner, S. A., & Gerloff, D. L. (1991) · Advances in Enzyme Regulation · doi:10.1016/0065-2571(91)90012-b · Cited by 170 · cited 1× · see citing works
“While traditional measures of covariations in natural sequences led to a few successes [1, 2, 3], major improvements came from recasting the problem as an inverse Potts model [4, 5].”
Background
Coevolutionary signals across protein lineages help capture multiple protein conformations
Morcos, F., Jana, B., Terence Hwa, B., & Onuchic, J. N. (2013) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1315625110 · Cited by 170 · cited 2× · see citing works
“Co-evolutionary patterns extracted from related sequences have proven useful to predict some IDR bound forms [151] and, in some cases, to untangle a protein’s multiple functional states [152, 153, 154].”
Background
High-resolution protein complexes from integrating genomic information with molecular simulation
Schug, A. H., Weigt, M., Onuchic, J. N., Terence Hwa, B., & Szurmant, H. (2009) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0912100106 · Cited by 162 · cited 1× · see citing works
“Such strategy has been employed to predict bacterial complexes [164, 165, 166], to single out pairs of interacting paralogs [167, 23] and to gain insight into the interactomes of viruses, like HCV [168], or bacteria, like E. coli [24].”
Background
ResPRE: high-accuracy protein contact prediction by coupling precision matrix with deep residual neural networks
Li, Y., Hu, J., Zhang, C., Yu, D.-J., & Zhang, Y. (2019) · Bioinformatics · doi:10.1093/bioinformatics/btz291 · Cited by 158 · cited 1× · see citing works
“More recently, some meta-models [87, 88], correlation and precision matrix-based approaches [89, 90, 52], and Overview of X-to-end and end-to-X deep learning approaches for protein structure prediction.”
Method
MoDEL (Molecular Dynamics Extended Library): A Database of Atomistic Molecular Dynamics Trajectories
Meyer, T., D’Abramo, M., Rueda, M., Ferrer‐Costa, C., Pérez, A., Carrillo, O., Camps, J., Fenollosa, C., Repchevsky, D., Gelpí, J. L., et al. (2010) · Structure · doi:10.1016/j.str.2010.07.013 · Cited by 157 · see citing works
Improved Contact Predictions Using the Recognition of Protein Like Contact Patterns
Skwark, M. J., Raimondi, D., Michel, M., & Elofsson, A. (2014) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1003889 · Cited by 155 · cited 1× · see citing works
“Shifting from unsupervised statistical inference to supervised deep learning further boosted the accuracy of the predicted contacts, and extended the applicability of this conceptual framework to families with fewer sequences [15, 16] and to the prediction of residue-residue distances [17, 18].”
BackgroundSelf-citation
GalaxyRefine2: simultaneous refinement of inaccurate local regions and overall protein structure
Lee, G. R., Won, J., Heo, L., & Seok, C. (2019) · Nucleic Acids Research · doi:10.1093/nar/gkz288 · Cited by 152 · see citing works
Fast end-to-end learning on protein surfaces
Sverrisson, F., Feydy, J., Correia, B. E., & Bronstein, M. M. (2021) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr46437.2021.01502 · Cited by 150 · cited 1× · see citing works
“Other developments include the application of ideas from the recent MoNet manifold network architecture(123) toward learning protein surfaces with very exciting outcomes for protein binding sites and protein–protein complexes prediction.(124,125) Overall, 2D surface manifolds seem very”
Background
Protein sequence design by conformational landscape optimization
Norn, C., Wicky, B. I. M., Juergens, D., Liu, S., Kim, D., Tischer, D., Koepnick, B., Anishchenko, I., Players, F., Baker, D., et al. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2017228118 · Cited by 144 · cited 1× · see citing works
“features takes place on the GPU and is tightly coupled to the network was implemented into trRosetta,(51) allowing for backpropagating gradients all the way to the input sequences.(56)”
Method
CONFOLD: Residue-residue contact-guidedab initioprotein folding
Adhikari, B., Bhattacharya, D., Cao, R., & Cheng, J. J. (2015) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.24829 · Cited by 143 · see citing works
ProteinNet: a standardized data set for machine learning of protein structure
AlQuraishi, M. (2019) · BMC Bioinformatics · doi:10.1186/s12859-019-2932-0 · Cited by 141 · see citing works
Ig-VAE: Generative Modeling of Protein Structure by Direct 3D Coordinate Generation
Eguchi, R. R., Choe, C. A., & Huang, P.-S. (2020) · doi:10.1101/2020.08.07.242347 · Cited by 141 · see citing works
Ig-VAE: Generative modeling of protein structure by direct 3D coordinate generation
Eguchi, R. R., Choe, C. A., & Huang, P.-S. (2022) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1010271 · Cited by 141 · see citing works
The Fitness Landscape of HIV-1 Gag: Advanced Modeling Approaches and Validation of Model Predictions by In Vitro Testing
Mann, J. K., Barton, J. P., Ferguson, A. L., Omarjee, S., Walker, B. D., Chakraborty, A. K., & Ndung’u, T. (2014) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1003776 · Cited by 136 · cited 2× · see citing works
“By leveraging the large amounts of available sequence data, several recent methods have achieved much higher accuracy than established popular approaches relying on a variety of sequence and structure-based features [173, 174, 175, 176, 177, 178, 179, 180, 181].”
BackgroundMethod
Correlated mutations in models of protein sequences: phylogenetic and structural effects
Lapedes, A. S., Giraud, B., Liu, L., & Stormo, G. D. (1999) · Institute of Mathematical Statistics Lecture Notes - Monograph Series · doi:10.1214/lnms/1215455556 · Cited by 124 · cited 2× · see citing works
“While traditional measures of covariations in natural sequences led to a few successes [1, 2, 3], major improvements came from recasting the problem as an inverse Potts model [4, 5].”
Background
Structured States of Disordered Proteins from Genomic Sequences
Tóth-Petróczy, Á., Palmedo, P., Ingraham, J., Hopf, T. A., Berger, B., Sander, C., & Marks, D. S. (2016) · Cell · doi:10.1016/j.cell.2016.09.010 · Cited by 124 · see citing works
GraphQA: protein model quality assessment using graph convolutional networks
Baldassarre, F., Hurtado, D. M., Elofsson, A., & Azizpour, H. (2020) · Bioinformatics · doi:10.1093/bioinformatics/btaa714 · Cited by 115 · see citing works
Self-citation
Large-scale discovery of protein interactions at residue resolution using co-evolution calculated from genomic sequences
Green, A. G., Elhabashy, H., Brock, K. P., Maddamsetti, R., Kohlbacher, O., & Marks, D. S. (2021) · Nature Communications · doi:10.1038/s41467-021-21636-z · Cited by 114 · cited 2× · see citing works
“These advances have significantly increased the protein structure modelling coverage of genomes [19, 20, 21], and also of bacterial inter-actomes [22, 23, 24].”
Background
Fitness landscape of the human immunodeficiency virus envelope protein that is targeted by antibodies
Louie, R. H. Y., Kaczorowski, K. J., Barton, J. P., Chakraborty, A. K., & McKay, M. R. (2018) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1717765115 · Cited by 111 · cited 2× · see citing works
“[175] Louie RH, Kaczorowski KJ, Barton JP, Chakraborty AK, McKay MR.”
BackgroundMethod
Protein contact prediction by integrating joint evolutionary coupling analysis and supervised learning
Ma, J., Wang, S., Wang, Z., & Xu, J. (2015) · Bioinformatics · doi:10.1093/bioinformatics/btv472 · Cited by 107 · cited 1× · see citing works
“Over the past 10 years, direct coupling analysis (DCA)-based models[12], including Potts model and pseudolikelihood maximization [8, 81, 82, 83], and Graphical lasso-based (low-rank) models [84, 85, 86] became widespread in the community.”
Method
Prediction of interresidue contacts with DeepMetaPSICOV in CASP13
Kandathil, S. M., Greener, J. G., & Jones, D. T. (2019) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.25779 · Cited by 107 · cited 1× · see citing works
“…"Recurrent Geometrical Network" comprising LSTM and geometric units and outputting a 3D structure a variety of of deep-learning models [91, 16, 92, 93, 21, 38, 73, 42, 41, 37, 45], including generative adversarial networks for contact map generation and reﬁne-ment [94, 35], got widely used to…”
Efficient generative modeling of protein sequences using simple autoregressive models
Trinquier, J., Uguzzoni, G., Pagnani, A., Zamponi, F., & Weigt, M. (2021) · Nature Communications · doi:10.1038/s41467-021-25756-4 · Cited by 106 · cited 1× · see citing works
“diction and protein design can be unified in a common theoretical framework extracting information from protein sequences.(171,172,175)”
Background
rawMSA: End-to-end Deep Learning using raw Multiple Sequence Alignments
Mirabello, C., & Wallner, B. (2019) · PLOS ONE · doi:10.1371/journal.pone.0220182 · Cited by 105 · cited 2× · see citing works
“One of the ﬁrst examples of end-to-X method was rawMSA [48], which leveraged embedding techniques from the ﬁeld of NLP, to map the amino acid residues into a continuous space adaptively learned based on the sequence context ( Table 1 ).”
Method
Prodock: Software package for protein modeling and docking
Trosset, J., & Scheraga, H. A. (1999) · Journal of Computational Chemistry · doi:10.1002/(sici)1096-987x(199903)20:4 · Cited by 103 · see citing works
Deep convolutional networks for quality assessment of protein folds
Derevyanko, G., Grudinin, S., Bengio, Y., & Lamoureux, G. (2018) · Bioinformatics · doi:10.1093/bioinformatics/bty494 · Cited by 101 · cited 2× · see citing works
“The ﬁrst attempt to train 3D CNNs on a volumetric protein representation dates back to CASP12, with the goal of assessing protein model quality [100].”
BackgroundSelf-citation
Protein model quality assessment using 3D oriented convolutional neural networks
Pagès, G., Charmettant, B., & Grudinin, S. (2019) · Bioinformatics · doi:10.1093/bioinformatics/btz122 · Cited by 94 · see citing works
Self-citation
Protein contact prediction using metagenome sequence data and residual neural networks
Wu, Q., Peng, Z., Anishchenko, I. V., Cong, Q., Baker, D. A., & Yang, J. (2019) · Bioinformatics · doi:10.1093/bioinformatics/btz477 · Cited by 86 · see citing works
Spherical harmonic molecular surfaces
Max, N., & Getzoff, E. D. (1988) · IEEE Computer Graphics and Applications · doi:10.1109/38.7748 · Cited by 85 · cited 1× · see citing works
“Spherical harmonics have played a prominent role in molecular surface representations for several decades [144, 145] and are also at the heart of the classical fast multipole method [146].”
Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks
(2021) · doi:10.1371/journal.pcbi.1008865 · Cited by 85 · cited 2× · see citing works
“…Network" comprising LSTM and geometric units and outputting a 3D structure a variety of of deep-learning models [91, 16, 92, 93, 21, 38, 73, 42, 41, 37, 45], including generative adversarial networks for contact map generation and reﬁne-ment [94, 35], got widely used to capture the same type of…”
Method
CopulaNet: Learning residue co-evolution directly from multiple sequence alignment for protein structure prediction
Ju, F., Zhu, J., Shao, B., Kong, L., Liu, T., Zheng, W., & Bu, D. (2021) · Nature Communications · doi:10.1038/s41467-021-22869-8 · Cited by 84 · see citing works
Dockground: A comprehensive data resource for modeling of protein complexes
Kundrotas, P. J., Anishchenko, I. V., Dauzhenka, T. A., Kotthoff, I., Mnevets, D., Copeland, M. M., & Vakser, I. A. (2017) · Protein Science · doi:10.1002/pro.3295 · Cited by 82 · cited 1× · see citing works
“In a small benchmark of 215 proteins [171], the structure for only a handful (5-10”
Background
Voronota: A fast and reliable tool for computing the vertices of the Voronoi diagram of atomic balls
Olechnovič, K., & Venclovas, Č. (2014) · Journal of Computational Chemistry · doi:10.1002/jcc.23538 · Cited by 80 · cited 1× · see citing works
“TheVoronoitessellationturnedout to be a very powerful description of protein structure and interactions and has been used in structural bioinformatics for several decades [112, 113, 114, 115, 116].”
Background
Revisiting the Voronoi description of protein–protein interfaces
Cazals, F., Martin, F. P.-D., Bahadur, R. P., & Janin, J. (2006) · Protein Science · doi:10.1110/ps.062245906 · Cited by 79 · see citing works
PconsC: combination of direct information methods and alignments improves contact prediction
Skwark, M. J., Abdel‐Rehim, A., & Elofsson, A. (2013) · Bioinformatics · doi:10.1093/bioinformatics/btt259 · Cited by 76 · cited 1× · see citing works
“More recently, some meta-models [87, 88], correlation and precision matrix-based approaches [89, 90, 52], and Overview of X-to-end and end-to-X deep learning approaches for protein structure prediction.”
MethodSelf-citation
3D Flexible Refinement: Structure and Motion of Flexible Proteins from Cryo-EM
Punjani, A., & Fleet, D. J. (2022) · Microscopy and Microanalysis · doi:10.1017/s1431927622005074 · Cited by 73 · see citing works
Smooth orientation-dependent scoring function for coarse-grained protein quality assessment
Karasikov, M., Pagès, G., & Grudinin, S. (2018) · Bioinformatics · doi:10.1093/bioinformatics/bty1037 · Cited by 69 · cited 1× · see citing works
“Historically, machine-learning-based scoring functions [135, 136] were inspired by statistical potentials [137, 138] relying on pairwise distance/ angular distributions or contact maps, which are perfectly rotation and translation invariant representations.”
BackgroundSelf-citation
VoroMQA web server for assessing three-dimensional structures of proteins and protein complexes
Olechnovič, K., & Venclovas, Č. (2019) · Nucleic Acids Research · doi:10.1093/nar/gkz367 · Cited by 69 · cited 1× · see citing works
“Another flavor of the VoroMQA method has also proved successful in scoring protein complexes since the CASP12-CAPRI experiment.(121,122) Here, only the protein–protein contact areas contribute”
Method
Accurate prediction of inter-protein residue–residue contacts for homo-oligomeric protein complexes
Yan, Y., & Huang, S. (2021) · Briefings in Bioinformatics · doi:10.1093/bib/bbab038 · Cited by 68 · see citing works
Local moves: An efficient algorithm for simulation of protein folding
Elofsson, A., Grand, S. M. L., & Eisenberg, D. S. (1995) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.340230109 · Cited by 64 · see citing works
Self-citation
Genomics and data science: an application within an umbrella
Navarro, F. C. P., Mohsen, H., Yan, C., Li, S., Gu, M., Meyerson, W. U., & Gerstein, M. (2019) · Genome biology · doi:10.1186/s13059-019-1724-1 · Cited by 62 · cited 1× · see citing works
“Genomics has become pre-eminent in terms of data scale, with an exponential growth.(64,65) These huge”
Background
Dilated Convolutions for Modeling Long-Distance Genomic DependenciesPreprint
Gupta, A., & Rush, A. M. (2017) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/200857 · Cited by 60 · cited 1× · see citing works
“With a dilation d , the window starting at location i of size Stacking dilated convolutions with increasingly large d allows operating on exponentially large receptive ﬁelds, while retaining short backpropagations [126, 127, 18].”
Background
Recent developments in deep learning applied to protein structure prediction
Kandathil, S. M., Greener, J. G., & Jones, D. T. (2019) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.25824 · Cited by 59 · cited 2× · see citing works
“This article is a follow-up to Kandathil et al.(32) It aims at providing CASP participants and observers with some overview of the recent developments in deep learning applied to protein structure prediction, and some comprehensive description of key concepts we think have”
Background
Hierarchical, rotation‐equivariant neural networks to select structural models of protein complexes
Eismann, S., Townshend, R. J. L., Thomas, N., Jagota, M., Jing, B., & Dror, R. O. (2020) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.26033 · Cited by 58 · see citing works
Self-citation
Fast end-to-end learning on protein surfacesPreprint
Sverrisson, F., Feydy, J., Correia, B. E., et al. (2020) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2020.12.28.424589 · Cited by 58 · see citing works
Residues Coevolution Guides the Systematic Identification of Alternative Functional Conformations in Proteins
Sfriso, P., Duran‐Frigola, M., Mosca, R., Emperador, A., Aloy, P., & Orozco, M. (2015) · Structure · doi:10.1016/j.str.2015.10.025 · Cited by 56 · cited 2× · see citing works
“[153] Sfriso P, Duran-Frigola M, Mosca R, Emperador A, Aloy P, Orozco M.”
Background
EigenTHREADER: analogous protein fold recognition by efficient contact map threading
Buchan, D. W. A., & Jones, D. T. (2017) · Bioinformatics · doi:10.1093/bioinformatics/btx217 · Cited by 56 · cited 1× · see citing works
“Both methods exploit predicted or observed inter-residue distances to improve the sequence alignments, a strategy that proved powerful already in CASP13.(72,79,80)”
Method
Inference of Epistatic Effects Leading to Entrenchment and Drug Resistance in HIV-1 Protease
Flynn, W. F., Haldane, A., Torbett, B. E., & Levy, R. M. (2017) · Molecular Biology and Evolution · doi:10.1093/molbev/msx095 · Cited by 55 · cited 2× · see citing works
“By leveraging the large amounts of available sequence data, several recent methods have achieved much higher accuracy than established popular approaches relying on a variety of sequence and structure-based features [173, 174, 175, 176, 177, 178, 179, 180, 181].”
BackgroundMethod
Fast and effective protein model refinement using deep graph neural networks
Jing, X., & Xu, J. (2021) · Nature Computational Science · doi:10.1038/s43588-021-00098-9 · Cited by 55 · cited 1× · see citing works
“interatomic potentials and employed them for protein structure refinement.(113) A more recent method, GVP-GNN,(114) augments graph networks with the ability to reason about protein features expressed”
Method
Low-N protein engineering with data-efficient deep learningPreprint
Biswas, S., Khimulya, G., Alley, E. C., et al. (2020) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2020.01.23.917682 · Cited by 54 · see citing works
Detecting distant-homology protein structures by aligning deep neural-network based contact maps
Zheng, W. H., Wuyun, Q., Li, Y., Mortuza, S. M., Zhang, C., Pearce, R., Ruan, J., & Zhang, Y. (2019) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1007411 · Cited by 51 · cited 1× · see citing works
“Both methods exploit predicted or observed inter-residue distances to improve the sequence alignments, a strategy that proved powerful already in CASP13.(72,79,80)”
Method
DEEPCON: protein contact prediction using dilated convolutional neural networks with dropout
Adhikari, B. (2019) · Bioinformatics · doi:10.1093/bioinformatics/btz593 · Cited by 49 · see citing works
Protein topology from predicted residue contacts
Taylor, W. R., Jones, D. T., & Sadowski, M. I. (2011) · Protein Science · doi:10.1002/pro.2002 · Cited by 46 · cited 1× · see citing works
“They enabled the modelling of 3D structures for large protein families [9, 10, 11, 12, 13, 14].”
Background
VoroCNN: deep convolutional neural network built on 3D Voronoi tessellation of protein structures
Igashov, I., Olechnovič, K., Kadukova, M., Venclovas, Č., & Grudinin, S. (2021) · Bioinformatics · doi:10.1093/bioinformatics/btab118 · Cited by 45 · see citing works
Self-citation
Trajectory-based training enables protein simulations with accurate folding and Boltzmann ensembles in cpu-hours
Jumper, J., Faruk, N. F., Freed, K. F., & Sosnick, T. R. (2018) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1006578 · Cited by 44 · cited 3× · see citing works
“So far, only a couple of learning methods have exploited solely and directly this information to efficiently fold proteins de novo [46, 47].”
Background
DeepECA: an end-to-end learning framework for protein contact prediction from a multiple sequence alignment
Fukuda, H., & Tomii, K. (2020) · BMC Bioinformatics · doi:10.1186/s12859-019-3190-x · Cited by 44 · cited 1× · see citing works
“More recently, some meta-models [87, 88], correlation and precision matrix-based approaches [89, 90, 52], and Overview of X-to-end and end-to-X deep learning approaches for protein structure prediction.”
Method
Fully differentiable full-atom protein backbone generation
Anand, N., Eguchi, R. R., & Huang, P. (2019) · International Conference on Learning Representations · Cited by 43 · see citing works
Improved fragment sampling for ab initio protein structure prediction using deep neural networks
Wang, T., Qiao, Y., Ding, W., Mao, W., Zhou, Y., & Gong, H. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-019-0075-7 · Cited by 42 · cited 3× · see citing works
“[63] Wang T, Qiao Y, DingW,MaoW, Zhou Y, GongH.”
Method
QDeep: distance-based protein model quality estimation by residue-level ensemble error classifications using stacked deep residual neural networks
Shuvo, M. H., Bhattacharya, S., & Bhattacharya, D. (2020) · Bioinformatics · doi:10.1093/bioinformatics/btaa455 · Cited by 41 · see citing works
Deep template-based protein structure prediction
Wu, F., & Xu, J. (2021) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1008954 · Cited by 41 · see citing works
A fully open-source framework for deep learning protein real-valued distances
Adhikari, B. (2020) · Scientific Reports · doi:10.1038/s41598-020-70181-0 · Cited by 40 · cited 3× · see citing works
“…Network" comprising LSTM and geometric units and outputting a 3D structure a variety of of deep-learning models [91, 16, 92, 93, 21, 38, 73, 42, 41, 37, 45], including generative adversarial networks for contact map generation and reﬁne-ment [94, 35], got widely used to capture the same type…”
Method
New scoring schemes for protein fold recognition based on Voronoi contacts.
Zimmer, R., W√∂hler, M., & Thiele, R. (1998) · Bioinformatics · doi:10.1093/bioinformatics/14.3.295 · Cited by 36 · cited 1× · see citing works
“TheVoronoitessellationturnedout to be a very powerful description of protein structure and interactions and has been used in structural bioinformatics for several decades [112, 113, 114, 115, 116].”
Background
Protein model accuracy estimation based on local structure quality assessment using 3D convolutional neural network
Sato, R., & Ishida, T. (2019) · PLoS ONE · doi:10.1371/journal.pone.0221347 · Cited by 35 · cited 6× · see citing works
“Sato-3DCNN by Sato and Ishida [104] used an idea similar to that of Ornate with oriented local frames but did not automatically learn the atom type embed-dings.”
BackgroundMethodInfluential
EPSILON-CP: using deep learning to combine information from multiple sources for protein contact prediction
Stahl, K., Schneider, M., & Brock, O. (2017) · BMC Bioinformatics · doi:10.1186/s12859-017-1713-x · Cited by 34 · cited 1× · see citing works
“…RaptorX [36], Galaxy, Triple-tRes [37], A2I2Prot, DESTINI2 [38], DeepHelicon [39], DeepHomo [40], ICOS, PrayogRealDistance [41, 42], RBO-PSP-CP [43], DeepECA, ropius0 [44], tFOLD, plus QUARK, Risoluto, Multicom [45] and those from the Zhang lab), several eﬀorts have been recently engaged…”
refineD: improved protein structure refinement using machine learning based restrained relaxation
Bhattacharya, D. (2019) · Bioinformatics · doi:10.1093/bioinformatics/btz101 · Cited by 34 · cited 1× · see citing works
“In GalaxyRefine2 [59], RefineD [60], and Baker suite [61], the QA is incorporated into a model refinement pipeline.”
Method
Phylogenetic correlations can suffice to infer protein partners from sequences
Marmier, G., Weigt, M., & Bitbol, A. (2019) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1007179 · Cited by 34 · cited 1× · see citing works
“Other methods trying to identify the pairs might work better but are computationally expensive.(166)”
Method
Interactome evolution: insights from genome-wide analyses of protein–protein interactions
Ghadie, M. A., Coulombe‐Huntington, J., & Xia, Y. (2017) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2017.10.012 · Cited by 32 · cited 1× · see citing works
“structure, one limitation here might be to detect a strong enough signal, since interactions across protein (and domain) interfaces are less conserved than intra-domain contacts.(168) Noticeably, the different”
Background
ProSPr: Democratized Implementation of Alphafold Protein Distance Prediction NetworkPreprint
Billings, W. M., Hedelius, B., Millecam, T., Wingate, D., & Corte, D. D. (2019) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/830273 · Cited by 32 · see citing works
GANcon: Protein Contact Map Prediction With Deep Generative Adversarial Network
Yang, H. X., Wang, M., Yu, Z., Zhao, X., & Li, A. (2020) · IEEE Access · doi:10.1109/access.2020.2991605 · Cited by 32 · cited 2× · see citing works
“[78] Yang H, Wang M, Yu Z, Zhao XM, Li A.”
Background
Single Layers of Attention Suffice to Predict Protein ContactsPreprint
Bhattacharya, N., Thomas, N., Rao, R., Dauparas, J., Koo, P. K., Baker, D. A., Song, Y. S., & Ovchinnikov, S. (2020) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2020.12.21.423882 · Cited by 31 · cited 1× · see citing works
“works have also shown that the residue-residue dependencies extracted by certain attention heads in transformers trained on large amounts of sequences can be directly interpreted as 3D contacts or distances.(48,133,134)”
Background
Coevolution analysis of Hepatitis C virus genome to identify the structural and functional dependency network of viral proteins
Champeimont, R., Laine, É., Hu, S., Pénin, F., & Carbone, A. (2016) · Scientific Reports · doi:10.1038/srep26401 · Cited by 30 · cited 1× · see citing works
“Such strategy has been employed to predict bacterial complexes [164, 165, 166], to single out pairs of interacting paralogs [167, 23] and to gain insight into the interactomes of viruses, like HCV [168], or bacteria, like E. coli [24].”
Background
Protein Contact Prediction from Amino Acid Co-Evolution Using Convolutional Networks for Graph-Valued Images
Golkov, V., Skwark, M. J., Golkov, A., Dosovitskiy, A., Brox, T., Meiler, J., & Cremers, D. (2016) · neural information processing systems · Cited by 29 · see citing works
Superadditive correlation
GIRAUD, B. G., Heumann, J. M., & Lapedes, A. S. (1999) · Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics · doi:10.1103/physreve.59.4983 · Cited by 29 · cited 1× · see citing works
“While traditional measures of co-variations in natural sequences led to a few successes, major improvements came from recasting the problem as an inverse Potts model.(4,5) These ideas started to show their full potential about 10 years ago with the development of efficient methods dealing with large scale multiple seq…”
Background
The role of coevolutionary signatures in protein interaction dynamics, complex inference, molecular recognition, and mutational landscapes
Morcos, F., & Onuchic, J. N. (2019) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2019.03.024 · Cited by 28 · cited 1× · see citing works
“Co-evolutionary patterns extracted from related sequences have proven useful to predict some IDR bound forms [151] and, in some cases, to untangle a protein’s multiple functional states [152, 153, 154].”
Background
Neural correlates of theory of mind in typically-developing youth: Influence of sex, age and callous-unemotional traits
Gao, Y., Rogers, J. C., Pauli, R., Clanton, R. L., Baker, R., Birch, P., Ferreira, L., Brown, A., Freitag, C. M., Fairchild, G. F., et al. (2019) · Scientific Reports · doi:10.1038/s41598-019-52261-y · Cited by 28 · see citing works
DeepDist: real-value inter-residue distance prediction with deep residual convolutional networkPreprint
Wu, T., Guo, Z., Hou, J., & Cheng, J. (2020) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2020.03.17.995910 · Cited by 25 · see citing works
Direct correlation analysis improves fold recognition
Sadowski, M. I., Maksimiak, K., & Taylor, W. R. (2011) · Computational Biology and Chemistry · doi:10.1016/j.compbiolchem.2011.08.002 · Cited by 22 · cited 1× · see citing works
“They enabled the modelling of 3D structures for large protein families [9, 10, 11, 12, 13, 14].”
Background
Spherical convolutions on molecular graphs for protein model quality assessment
Igashov, I., Pavlichenko, N., & Grudinin, S. (2021) · Machine Learning Science and Technology · doi:10.1088/2632-2153/abf856 · Cited by 21 · see citing works
Self-citation
Prediction of inter-residue contacts with DeepMetaPSICOV in CASP13Preprint
Kandathil, S. M., Greener, J. G., & Jones, D. T. (2019) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/586800 · Cited by 20 · see citing works
Modeling of protein complexes in CASP14 with emphasis on the interaction interface prediction
Dapkūnas, J., Olechnovič, K., & Venclovas, Č. (2021) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.26167 · Cited by 19 · see citing works
Modeling conformational flexibility of kinases in inactive states
Schwarz, D., Merget, B., Deane, C. M., & Fulle, S. (2019) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.25756 · Cited by 18 · cited 1× · see citing works
“For example, the inactive state of kinases is largely underrepresented in the PDB compared to the active state.(155) This may bias data-driven approaches while, in”
Background
A novel sequence alignment algorithm based on deep learning of the protein folding code
Gao, M., & Skolnick, J. (2021) · Bioinformatics · doi:10.1093/bioinformatics/btaa810 · Cited by 18 · see citing works
ProALIGN: Directly Learning Alignments for Protein Structure Prediction via Exploiting Context-Specific Alignment Motifs
Kong, L., Ju, F., Zheng, W., Zhu, J., Sun, S., Xu, J., & Bu, D. (2022) · Journal of Computational Biology · doi:10.1089/cmb.2021.0430 · Cited by 16 · cited 3× · see citing works
“NDThreader [77] and ProALIGN [78] are speciﬁcally designed to optimally align the query with the template in template-based modeling.”
BackgroundInfluential
PconsFam: An Interactive Database of Structure Predictions of Pfam Families
Lamb, J., Jarmolinska, A. I., Michel, M., Hurtado, D. M., Sulkowska, J. I., & Elofsson, A. (2019) · Journal of Molecular Biology · doi:10.1016/j.jmb.2019.01.047 · Cited by 15 · cited 1× · see citing works
“In a small benchmark of 215 proteins,(170) the structure for only a handful (5%–10%) of the complexes could be predicted correctly using a naive approach matching top hits from all genomes.(20) Other methods trying to identify the pairs might work better but are computationally expensive.”
MethodSelf-citation
DeepHelicon: Accurate prediction of inter-helical residue contacts in transmembrane proteins by residual neural networks
Sun, J., & Frishman, D. (2020) · Journal of Structural Biology · doi:10.1016/j.jsb.2020.107574 · Cited by 14 · cited 1× · see citing works
“…ProSPr [34], Kiharalab_Contact [35], Phar-mulator, DeepPotential, RaptorX [36], Galaxy, Triple-tRes [37], A2I2Prot, DESTINI2 [38], DeepHelicon [39], DeepHomo [40], ICOS, PrayogRealDistance [41, 42], RBO-PSP-CP [43], DeepECA, ropius0 [44], tFOLD, plus QUARK, Risoluto, Multicom [45] and those…”
Structural modeling of protein complexes: Current capabilities and challenges
Dapkūnas, J., Olechnovič, K., & Venclovas, Č. (2019) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.25774 · Cited by 13 · see citing works
Large-scale clinical interpretation of genetic variants using evolutionary data and deep learningPreprint
Frazer, J., Notin, P., Dias, M., et al. (2020) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2020.12.21.423785 · Cited by 13 · cited 1× · see citing works
“These approaches make the estimation of the impact of every possible substitution at every position in a proteincoding genome computationally feasible.(181) They also hold great potential for guiding protein design and engineering.”
Background
Graphical models of residue coupling in protein families
Thomas, J. M., Ramakrishnan, N., & Bailey‐Kellogg, C. (2005) · doi:10.1145/1134030.1134033 · Cited by 11 · cited 1× · see citing works
“These ideas started to show their full potential about 10 years ago with the development of eﬃcient methods dealing with large scale multiple sequence alignments [6, 7, 8].”
Background
Protein-protein docking using learned three-dimensional representationsPreprint
Derevyanko, G., & Lamoureux, G. (2019) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/738690 · Cited by 9 · cited 3× · see citing works
“…atom embeddings Sato-3DCNN[104] A local frame-based 3D CNN 3DCNN_prof (P3CMQA)[105] Extends Sato-3DCNN with predicted features and PSSMs SE(3)-3DCNN[101] An invariant 3D CNN based on [102] trained for protein complexes iPhord & DeepMUSICS 3D CNNs TopQA[125] 3D CNN with explicit rotations and…”
BackgroundMethodInfluential
TopQA: a topological representation for single-model protein quality assessment with machine learning
Smith, J., Conover, M., Stephenson, N., Eickholt, J., Si, D., Sun, M., & Cao, R. (2020) · International Journal of Computational Biology and Drug Design · doi:10.1504/ijcbdd.2020.105095 · Cited by 9 · see citing works
COMER2: GPU-accelerated sensitive and specific homology searches
Margelevičius, M. (2020) · Bioinformatics · doi:10.1093/bioinformatics/btaa185 · Cited by 9 · cited 1× · see citing works
“1, see iPhord, ProSPr [34], Kiharalab_Contact [35], Pharmulator, DeepPotential, RaptorX [36], Galaxy, TripletRes [37], A2I2Prot, DESTINI2 [38], DeepHelicon [39], DeepHomo [40], ICOS, PrayogRealDistance [41, 42], RBO-PSP-CP [43], DeepECA, ropius0 [44], tFOLD, plus QUARK, Risoluto, Multicom [45] and those from the Zhang…”
Background
P3CMQA: Single-Model Quality Assessment Using 3DCNN with Profile-Based Features
Takei, Y., & Ishida, T. (2021) · Bioengineering · doi:10.3390/bioengineering8030040 · Cited by 8 · see citing works
Ultrafast end-to-end protein structure prediction enables high-throughput exploration of uncharacterised proteins
Kandathil, S. M., Greener, J. G., Lau, A. M., & Jones, D. T. (2020) · doi:10.1101/2020.11.27.401232 · Cited by 8 · cited 9× · see citing works
“The most recent version of DMPfold2 (DPMfold2-new in Table 1 [50]) attempted to resolve this issue by adding an extra-GRU layer.”
BackgroundMethodInfluential
Protein model quality assessment using rotation-equivariant, hierarchical neural networksPreprint
Eismann, S., Suriana, P., Jing, B., Townshend, R. J. L., & Dror, R. O. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.13557 · Cited by 7 · cited 7× · see citing works
“[95], starting from the 3D coordinates and element type of each atom, the network first learns rotation-equivariant representations of local neighborhoods and then aggregates this information hierarchically to predict a rotation-invariant fingerprint at the level of the entire protein structure, reflecting the previou…”
BackgroundMethodInfluentialSelf-citation
ISSEC: inferring contacts among protein secondary structure elements using deep object detection
Zhang, Q., Zhu, J., Ju, F., Kong, L., Sun, S., Zheng, W., & Bu, D. (2020) · BMC Bioinformatics · doi:10.1186/s12859-020-03793-y · Cited by 4 · see citing works
Untitled work
(1999) · doi:10.1002/(sici)1097-0134(1999)37:3+

Flagged first, then by citations

Research works printed but not linked: 22

[27][27] Cohen TS, Welling M. Steerable cnns. arXiv preprint arXiv:161208498 2016;. -
[28][28] Gilmer J, Schoenholz SS, Riley PF, Vinyals O, Dahl GE. Neural Message Passing for Quantum Chemistry. In: Precup D, Teh YW, editors. Proceedings of the 34th International Conference on Machine Learning, vol. 70 of Proceedings of Machine Learning Research PMLR; 2017. p. 1263-1272. http://proceedings. mlr.press/v /gilmer a.html . -
[33][33] Elofsson A, LeGrand S, Eisenberg D. Local moves, an efficient method for protein folding simulations. Proteins 1995;23:73-82. -
[35][35] Subramaniya SRMV, Terashi G, Jain A, Kagaya Y, Kihara D. Protein contact map denoising using generative adversarial networks. bioRxiv 2020;. -
[38][38] GaoM,ZhouH,SkolnickJ. DESTINI: A deep-learning approach to contact-driven protein structure prediction. Scientific reports 2019;9(1):1-13. -
[40][40] Yan Y, Huang SY. Accurate prediction of residueresidue contacts across homo-oligomeric protein interfaces through deep leaning. bioRxiv 2020;. -
[47][47] Ingraham J, Riesselman A, Sander C, Marks D. Learning protein structure with a differentiable simulator. In: 7th International Conference on Learning Representations, ICLR New Orleans, LA, USA, May 6-9, 2019; 2019. . -
[49][49] Kandathil SM, Greener JG, Lau AM, Jones DT. Deep learning-based prediction of protein structure using learned representations of multiple sequence alignments. bioRxiv 2020;. -
[56][56] Adhikari B, Bhattacharya D, Cao R, Cheng J. CONFOLD: Residue-residue contact-guided ab initio protein folding. Proteins 2015 Aug;83(8):1436-1449. -
[55][55] Drori I, Thaker D, Srivatsa A, Jeong D, Wang Y, Nan L, et al. Accurate protein structure prediction by embeddings and deep learning representations. arXiv preprint arXiv:191105531 2019;. -
[96][96] Peters ME, Neumann M, Iyyer M, Gardner M, Clark C, Lee K, et al. Deep contextualized word representations. arXiv preprint arXiv:180205365 2018;. -
[97][97] Radford A, Wu J, Child R, Luan D, Amodei D, Sutskever I. Language models are unsupervised multitask learners. OpenAI blog 2019;1(8):9. -
[110][110] Eismann S, Townshend RJ, Thomas N, Jagota M, Jing B, Dror R. Hierarchical, rotation-equivariant neural networks to predict the structure of protein complexes. arXiv preprint arXiv:200609275 2020;. -
[109][109] Jing B, Eismann S, Suriana P, Townshend RJL, Dror R. Learning from Protein Structure with Geometric Vector Perceptrons. In: International Conference on Learning Representations; 2021. https:// openreview.net/forum?id= YLJDvSx J . -
[132][132] Veličković P, Cucurull G, Casanova A, Romero A, Liò P, Bengio Y. Graph Attention Networks. International Conference on Learning Representations 2018; https://openreview.net/forum?id= rJXMpikCZ . -
[139][139] Kondor R. N-body networks: a covariant hierarchical neural network architecture for learning atomic potentials. arXiv preprint arXiv:180301588 2018;. -
[140][140] Anderson B, Hy TS, Kondor R. Cormorant: Covariant molecular neural networks. arXiv preprint arXiv:190604015 2019;. -
[142][142] Schütt KT, Kindermans PJ, Sauceda HE, Chmiela S, Tkatchenko A, Müller KR. SchNet: A continuousfilter convolutional neural network for modeling quantum interactions. arXiv:170608566 2017;. -
[145][145] Ritchie DW, Kemp GJ. Fast computation, rotation, and comparison of low resolution spherical harmonic molecular surfaces. Journal of Computational Chemistry 1999;20(4):383-395. -
[158][158] Rosenbaum D, Garnelo M, Zielinski M, Beattie C, Clancy E, Huber A, et al. Inferring a Continuous Distribution of Atom Coordinates from Cryo-EM Images using VAEs. arXiv preprint arXiv:210614108 2021;. -
[170][170] Quadir F, Roy R, Halfmann R, Cheng J. Predicting interchain contacts for homodimeric and homomultimeric protein complexes using multiple sequence alignments of monomers and deep learning. bioRxiv 2020; https://www.biorxiv.org/ content/early/ / / / . . . . -
[172][172] Weinstein EN, Marks DS. A structured observation distribution for generative biological sequence prediction and forecasting. bioRxiv 2021; https://www.biorxiv.org/content/early/ / / / . . . . -

Sorted by citations

Critical assessment of methods of protein structure prediction (CASP)—Round <scp>XIV</scp>
Kryshtafovych, A., Schwede, T., Topf, M., Fidelis, K., & Moult, J. (2021) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.26237 · Cited by 503 · see citing works
Protein Design: From the Aspect of Water Solubility and Stability
Qing, R., Hao, S.-L., Smorodina, E., Jin, D., Zalevsky, A. O., & Zhang, S.-G. (2022) · Chemical Reviews · doi:10.1021/acs.chemrev.1c00757 · Cited by 287 · see citing works
PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces
Krapp, L. F., Abriata, L. A., Rodríguez, F. C., & Peraro, M. D. (2023) · Nature Communications · doi:10.1038/s41467-023-37701-8 · Cited by 159 · see citing works
“The major breakthroughs come from the ﬁ eld of protein folding 30 , where AlphaFold 19 integrates attention in the Evoformer blocks and the structure module and the third track of the RoseTTAFold 20 model uses a SE(3)- Transformer 31 to re ﬁ ne the atom coordinates during folding.”
Progress at protein structure prediction, as seen in CASP15
Elofsson, A. (2023) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2023.102594 · Cited by 89 · see citing works
Self-citation
On the Potential of Machine Learning to Examine the Relationship Between Sequence, Structure, Dynamics and Function of Intrinsically Disordered Proteins
Lindorff-Larsen, K., & Kragelund, B. B. (2021) · Journal of Molecular Biology · doi:10.1016/j.jmb.2021.167196 · Cited by 85 · see citing works
Deep learning for protein structure prediction and design—progress and applications
Jänes, J., & Beltrão, P. (2024) · Molecular Systems Biology · doi:10.1038/s44320-024-00016-x · Cited by 71 · see citing works
“These advances and their historical perspective have been reviewed elsewhere (AlQuraishi, 2021; Laine et al, 2021; Elofsson, 2023) and can be summarized by an increase in usage of neural network models along key parts of the protein structure prediction problem.”
Background
The Transporter-Mediated Cellular Uptake and Efflux of Pharmaceutical Drugs and Biotechnology Products: How and Why Phospholipid Bilayer Transport Is Negligible in Real Biomembranes
Kell, D. B. (2021) · Molecules · doi:10.3390/molecules26185629 · Cited by 48 · see citing works
Protein Design with Deep Learning
Defresne, M., Barbe, S., & Schiex, T. (2021) · International Journal of Molecular Sciences · doi:10.3390/ijms222111741 · Cited by 44 · see citing works
“Indeed, protein structures are naturally insensitive to translations and rotations; locally, only the relative orientation and position of structural motifs matters [11].”
Equivariant Graph Neural Networks for 3D Macromolecular StructurePreprint
Jing, B., Eismann, S., Soni, P. N., & Dror, R. O. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.03843 · Cited by 43 · see citing works
Self-citation
Accurate prediction of protein assembly structure by combining AlphaFold and symmetrical docking
Jeppesen, M., & André, I. (2023) · Nature Communications · doi:10.1038/s41467-023-43681-6 · Cited by 44 · see citing works
DeepHomo2.0: improved protein–protein contact prediction of homodimers by transformer-enhanced deep learning
Lin, P., Yan, Y., & Huang, S. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac499 · Cited by 39 · see citing works
Can AlphaFold’s breakthrough in protein structure help decode the fundamental principles of adaptive cellular immunity?
McMaster, B., Thorpe, C. J., Ogg, G. S., Deane, C. M., & Koohy, H. (2024) · Nature Methods · doi:10.1038/s41592-024-02240-7 · Cited by 41 · see citing works
“Since deep-learning models made waves in the CASP14 competition 55 , new methods have rapidly flooded the field, offering the potential for fresh insights into TCR–pMHC interactions.”
Structural Bioinformatics and Deep Learning of Metalloproteins: Recent Advances and Applications
Andreini, C., & Rosato, A. (2022) · International Journal of Molecular Sciences · doi:10.3390/ijms23147684 · Cited by 25 · see citing works
Beyond AlphaFold2: The Impact of AI for the Further Improvement of Protein Structure Prediction
Genc, A. G., & McGuffin, L. J. (2024) · Methods in molecular biology · doi:10.1007/978-1-0716-4196-5_7 · Cited by 22 · see citing works
Machine learning in computational modelling of membrane protein sequences and structures: From methodologies to applications
Sun, J., Kulandaisamy, A., Liu, J., Hu, K., Gromiha, M. M., & Zhang, Y. (2023) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2023.01.036 · Cited by 19 · see citing works
“In particular, the frontier of 3D structural prediction has recently witnessed a rapid transition from conventional machine learning to deep learning methods [110,111].”
Background
Alignment-based Protein Mutational Landscape Prediction: Doing More with Less
Abakarova, M., Marquet, C., Rera, M., Rost, B., & Laine, E. (2023) · Genome Biology and Evolution · doi:10.1093/gbe/evad201 · Cited by 14 · see citing works
“In recent years, protein 3D structures and mutations’ functional impact prediction (Laine et al. 2021; Nature Methods 2022) has achieved tremendous progress by leveraging the wealth of publicly available natural protein sequence data (Nordberg et al. 2014; Suzek et al. 2015; Steinegger and Söding…”
Background
Using metagenomic data to boost protein structure prediction and discovery
Hou, Q., Pucci, F., Pan, F., Xue, F.-Z., Rooman, M., & Feng, Q. (2022) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2021.12.030 · Cited by 16 · see citing works
“the predictors [8], which have achieved astonishing scores in recent years [9,10].”
Background
A STRP-ed definition of Structured Tandem Repeats in Proteins
Monzón, A. M., Arrías, P. N., Elofsson, A., Mier, P., Andrade-Navarro, M. A., Bevilacqua, M., Clementel, D., Bateman, A. G., Hirsh, L., Fornasari, M. S., et al. (2023) · Journal of Structural Biology · doi:10.1016/j.jsb.2023.108023 · Cited by 14 · see citing works
Self-citation
Cryptic Oxidative Transamination of Hydroxynaphthoquinone in Natural Product Biosynthesis
Noguchi, T., Isogai, S., Terada, T., Nishiyama, M., & Kuzuyama, T. (2022) · Journal of the American Chemical Society · doi:10.1021/jacs.1c13074 · Cited by 11 · see citing works
Machine learning approaches for predicting protein-ligand binding sites from sequence data
Vural, O., & Jololian, L. (2025) · Frontiers in Bioinformatics · doi:10.3389/fbinf.2025.1520382 · Cited by 8 · see citing works
Navigating protein–nucleic acid sequence-structure landscapes with deep learning
Laine, É., Grudinin, S., Klypa, R., & de Beauchêne, I. C. (2025) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2025.103162 · Cited by 6 · see citing works
“The spectacular advances in protein structure prediction [6,7] have stimulated a strong interest in expanding successful deep learning architectures to model nucleic acids in addition to proteins (Figure 1a,c and Table 1).”
Self-citation
Prospective on methods of design of experiments for limited data scenarios in materials design and engineering
Ryan, E. M., Roshandelpoor, A., Pollard, Z. A., Goldfarb, J. L., & Vakili, P. (2023) · MRS Communications · doi:10.1557/s43579-023-00478-4 · Cited by 6 · see citing works
Quantum Computing and Its Promise in Drug Discovery
Mishra, R., Mishra, P. S., Mazumder, R., Mazumder, A., & Varshney, S. (2024) · doi:10.1002/9781394159338.ch3 · Cited by 6 · see citing works
How much metagenome data is needed for protein structure prediction: The advantages of targeted approach from the ecological and evolutionary perspectives
Yang, P., & Ning, K. (2022) · iMeta · doi:10.1002/imt2.9 · Cited by 4 · see citing works
Solvent Accessibility Promotes Rotamer Errors during Protein Modeling with Major Side-Chain Prediction Programs
Hameduh, T., Mokry, M., Miller, A. D., Heger, Z., & Haddad, Y. (2023) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.3c00134 · Cited by 5 · see citing works
AlphaFold Database Structure Extractor: a web server and API to download AlphaFold structures using common protein accessions
Saraf, N., Karthik, V., & Sharma, G. (2025) · BMC Bioinformatics · doi:10.1186/s12859-025-06303-0 · Cited by 4 · see citing works
Enhancing coevolutionary signals in protein–protein interaction prediction through clade-wise alignment integration
Fang, T., Szklarczyk, D., Hachilif, R., & von Mering, C. (2024) · Scientific Reports · doi:10.1038/s41598-024-55655-9 · Cited by 4 · see citing works
“Many computational methods could be further applied to extract direct physical interactions 19,26 .”
Method
Metal-induced structural variability of mononuclear metal-binding sites from a database perspective
Bazayeva, M., Laveglia, V., Andreini, C., & Rosato, A. (2022) · Journal of Inorganic Biochemistry · doi:10.1016/j.jinorgbio.2022.112025 · Cited by 4 · see citing works
Improving Inter-Helix Contact Prediction With Local 2D Topological Information
Li, J., Sawhney, A., Lee, J., & Liao, L. (2023) · IEEE Transactions on Computational Biology and Bioinformatics · doi:10.1109/tcbb.2023.3274361 · Cited by 4 · see citing works
6DCNN with Roto-Translational Convolution Filters for Volumetric Data Processing
Zhemchuzhnikov, D., Igashov, I., & Grudinin, S. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i4.20396 · Cited by 4 · see citing works
Self-citation
Alignment-based protein mutational landscape prediction: doing more with less
(2022) · doi:10.1101/2022.12.13.520259 · Cited by 4 · see citing works
xTrimoDock: Rigid Protein Docking via Cross-Modal Representation Learning and Spectral AlgorithmPreprint
Luo, Y., Li, S., Sun, Y., Wang, R., Tang, T., Hongdu, B., Cheng, X., Shi, C., Li, H., & Song, L. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.02.06.527251 · Cited by 4 · see citing works
A Generative Neuro‐Symbolic AI for Protein Sequence Design
Defresne, M., Dessaux, D., Buchet, S., Barthe, L., Khodja, L. A., Azizi, B., Durante, V., Cioci, G., de Givry, S., Roussel, A., et al. (2026) · Advanced Science · doi:10.1002/advs.76464 · Cited by 1 · see citing works
Enhancing the Performance of AlphaFold Through Modified Storage Method and Optimization of HHblits on TSUBAME3.0 Supercomputer
Fujita, H., Nomura, A., Endo, T., & Sekijima, M. (2023) · Congress in Computer Science, Computer Engineering, and Applied Computing (CSCE) · doi:10.1109/csce60160.2023.00351 · Cited by 1 · see citing works
“Under such a background, AlphaFold v2 [9] emerged in the 14th Critical Assessment of Structure Prediction (CASP14) [10], [11] and achieved results that greatly exceeded those of other methods.”
Result
PeSTo: parameter-free geometric deep learning for accurate prediction of protein interacting interfacesPreprint
Krapp, L. F., Abriata, L. A., Rodríguez, F. C., & Peraro, M. D. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.05.09.491165
“The choice of a representation for a specific problem is important as it will affect the overall performance of a machine learning model [21].”
Background
Metal-Induced Structural Variability of Metal-Binding Sites from a Database PerspectivePreprint
Bazayeva, M., Laveglia, V., Andreini, C., & Rosato, A. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4182153
Clade-wise alignment integration improves co-evolutionary signals for protein-protein interaction predictionPreprint
Fang, T., Szklarczyk, D., Hachilif, R., & von Mering, C. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.07.28.550005
“Many computational methods could be further applied to extract direct physical interactions 19,26 .”
Method
AutoRNC: An automated modeling program for building atomic models of ribosome-nascent chain complexes
(2024) · doi:10.1016/j.str.2024.02.002
Clustering Diffusion Model for Improving Protein Structure Prediction Performance
Zhong, W., Zhang, X., & Yu, N. (2025) · International Conference on Artificial Intelligence and Big Data (ICAIBD) · doi:10.1109/icaibd64986.2025.11082075
“The success of deep learning-based AI models depends on a large number of samples in the training set [7].”
Transformer-based classification and interpretability of NR3C1 expression patterns in oral squamous cell carcinoma: Metabolic adaptation insights
Yuwanati, M. B., Yadalam, P. K., & Mullainathan, S. M. (2026) · Journal of Oral Biology and Craniofacial Research · doi:10.1016/j.jobcr.2026.101427
New computational approaches to study the effect of genetic mutations on the topology of protein-protein interaction networks
Behbahani, Y. M. (2022) · theses.fr (ABES) · doi:10.70675/bfdded7czb638z455azaedaz4ae9f8d0a13d
Structural prediction of protein assemblies with deep learning
Bret, H. (2023) · theses.fr (ABES) · doi:10.70675/b3849091zce01z4ab0z91efz2b71d9cc754d
Data science approach for the exploration of HLA antigenicity based on 3D structures and molecular dynamics
Amaya-Ramírez, D. (2024) · doi:10.70675/00312c21z4791z4a88zabfez5e30cf36e0df
Structural compatibility is not enough: Current challenges in protein-protein interaction prediction
Elofsson, A. (2026) · Molecular Cell · doi:10.1016/j.molcel.2026.08.030
Self-citation

Deep learning entering the post-protein structure prediction era: new horizons for structural biology
Grudinin, S. (2021) · Acta Crystallographica Section A Foundations and Advances · doi:10.1107/s0108767321092138
Self-citation
Application of deep learning methods for protein structure prediction
Golenko, Y., Ismailova, A., & Moldasheva, R. N. (2022) · Bulletin of the National Engineering Academy of the Republic of Kazakhstan · doi:10.47533/2020.1606-146x.192
Deep Learning-Based Prediction of Protein Structure from Sequence Data
Dubois, P., Havlicek, C., & Fernandez, A. (2024) · Zenodo (CERN European Organization for Nuclear Research) · doi:10.5281/zenodo.22688030
Deep Learning Models for Protein Structure Prediction: A Comprehensive Survey
Srushti, C., Prathibhavani, P. M., & Venugopal, K. R. (2026) · International Journal of Emerging Research in Engineering Science and Management · doi:10.58482/ijeresm.v5i2.7
Recent Applications of Deep Learning Methods on Evolution- and Contact-Based Protein Structure Prediction
Suh, D., Lee, J. W., Choi, S., & Lee, Y. (2021) · International Journal of Molecular Sciences · doi:10.3390/ijms22116032 · Cited by 25 · see citing works

Authors: Laine, E., Eismann, S., Elofsson, A., & Grudinin, S.
Affiliations: Sorbonne Université, CNRS, IBPS, Laboratoire de Biologie Computationnelle et Quantitative (LCQB) Paris France
Stanford University (US)
Stockholm University (SE)
Institut polytechnique de Grenoble (FR)
ORCID: Laine, Elodie
Elofsson, Arne
Grudinin, Sergei
Venue: Proteins: Structure, Function, and Bioinformatics
Publisher: Wiley
Published: 2021-09-22
Volume: 89
Issue: 12
Pages: 1770-1786
ISSN: 0887-3585 (ISSN-L)
1097-0134
DOI: 10.1002/prot.26235
OpenAlex: W3160144628
Cited by: 44
References: 173
Citations by year: 2026: 3 · 2025: 4 · 2024: 7 · 2023: 13 · 2022: 12 · 2021: 5
FWCI: 0.97
Type: journal article
Language: en
Open access: green
arxiv.org
Licence: onlinelibrary.wiley.com/termsAndConditions#vor (vor)
Locations: Proteins Structure Function and Bioinformatics (publishedVersion)
PubMed (publishedVersion)
arXiv (Cornell University) (submittedVersion)
HAL (submittedVersion, other-oa)
arXiv (Cornell University)
Funders: Agence Nationale de la Recherche (ANR‐17‐CE12‐0009)
National Institute for Research in Computer Science and Control
Swedish e-Science Research Centre
Topics: Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
Physical Sciences › Materials Science › Materials Chemistry › Enzyme Structure and Function
Keywords: protein structure prediction, end-to-end architecture, geometric deep learning, graph neural networks, Voronoi tessellation, point clouds, attention mechanism, protein representation, CASP14
MeSH terms: Deep Learning, Amino Acid Sequence, Protein Conformation, Proteins, Software, Computational Biology, Sequence Analysis, Protein, Databases, Protein

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Protein sequence‐to‐structure learning: Is this the end(‐to‐end revolution)?

Abstract

Integrity signals

Claims the paper makes

Claims other papers make about this one

Self-check

Cites

Cited by

Related

Record

About these checks