Protein–protein interaction prediction with deep learning: A comprehensive review
Open accessThe DOI matches its Crossref record; 64 of the 64 references it cites were found in a registry or an index.
Abstract
Most proteins perform their biological function by interacting with themselves or other molecules. Thus, one may obtain biological insights into protein functions, disease prevalence, and therapy development by identifying protein-protein interactions (PPI). However, finding the interacting and non-interacting protein pairs through experimental approaches is labour-intensive and time-consuming, owing to the variety of proteins. Hence, protein-protein interaction and protein-ligand binding problems have drawn attention in the fields of bioinformatics and computer-aided drug discovery. Deep learning methods paved the way for scientists to predict the 3-D structure of proteins from genomes, predict the functions and attributes of a protein, and modify and design new proteins to provide desired functions. This review focuses on recent deep learning methods applied to problems including predicting protein functions, protein-protein interaction and their sites, protein-ligand binding, and protein design.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Szklarczyk, D., Gable, A. L., Lyon, D., Junge, A., Wyder, S., Huerta-Cepas, J., Simonovic, M., Doncheva, N. T., Morris, J. H., Bork, P., et al. (2019) · Nucleic Acids Research · doi:10.1093/nar/gky1131 · Cited by 17,749
Unsupervised representation learning with deep convolutional generative adversarial networks
Radford, A., Metz, L., & Chintala, S. (2015) · arXiv · doi:10.48550/arxiv.1511.06434 · Cited by 12,235
Cd-hit: Accelerated for clustering the next-generation sequencing data
Fu, L., Niu, B., Zhu, Z., Wu, S., & Li, W. (2012) · Bioinformatics · doi:10.1093/bioinformatics/bts565 · Cited by 11,175
A Comprehensive Survey on Graph Neural Networks
Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., & Yu, P. S. (2021) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2020.2978386 · Cited by 11,000 · cited 1×
“However, when data cannot be represented on a regular grid due to the complex nature of their correlations [135,136], standard convolution cannot be directly applied to non-Euclidean geometries, limiting the applicability of CNNs [137,138].”
BackgroundLearning long-term dependencies with gradient descent is difficult
Bengio, Y., Simard, P., & Frasconi, P. (1994) · IEEE Transactions on Neural Networks · doi:10.1109/72.279181 · Cited by 8,542 · cited 1×
“However, RNNs often suffer from vanishing gradients, limiting the context range they can access [128,129].”
BackgroundSzklarczyk, D., Gable, A. L., Nastou, K. C., Lyon, D., Kirsch, R., Pyysalo, S., Doncheva, N. T., Legeay, M., Fang, T., Bork, P., et al. (2021) · Nucleic Acids Research · doi:10.1093/nar/gkaa1074 · Cited by 8,056 · cited 12×
“STRING, IntAct and MINT provide interaction scores from different sources to indicate their reliability.”
BackgroundMethodInfluentialUniProt: a worldwide hub of protein knowledge
The UniProt Consortium (2019) · Nucleic Acids Research · doi:10.1093/nar/gky1049 · Cited by 7,698 · cited 3×
“Protein sequences can be obtained from the STRING [109], PDB [110], UniProt [106], PIR [119], SWISS-MODEL [107], and TrEMBL [120] databases.”
MethodXGBoost: A Scalable Tree Boosting System
Chen, T., & Guestrin, C. (2016) · arXiv · doi:10.48550/arxiv.1603.02754 · Cited by 7,104
Principles that Govern the Folding of Protein Chains
Anfinsen, C. B. (1973) · Science · doi:10.1126/science.181.4096.223 · Cited by 7,096
Szklarczyk, D., Morris, J. H., Cook, H., Kuhn, M., Wyder, S., Simonovic, M., Santos, A., Doncheva, N. T., Roth, A., Bork, P., et al. (2017) · Nucleic Acids Research · doi:10.1093/nar/gkw937 · Cited by 6,980
UniProt: the universal protein knowledgebase in 2021
The UniProt Consortium, Bateman, A., Martin, M.-J., Orchard, S., Magrane, M., Agivetova, R., Ahmad, S., Alpi, E., Bowler-Barnett, E. H., Britto, R., et al. (2021) · Nucleic Acids Research · doi:10.1093/nar/gkaa1100 · Cited by 6,372 · cited 7×
“Protein sequences were obtained from the UniProt [327] and restricted to lengths of 150–1,500 amino acids.”
BackgroundMethodInfluentialShuman, D. I., Narang, S. K., Frossard, P., Ortega, A., & Vandergheynst, P. (2013) · IEEE Signal Processing Magazine · doi:10.1109/msp.2012.2235192 · Cited by 4,421 · cited 8×
“However, the convolution theorem states that [137] convolution may be evaluated using Fourier transform.”
BackgroundMethodInfluentialStarGAN: Unified Generative Adversarial Networks for Multi-domain Image-to-Image Translation
Choi, Y., Choi, M., Kim, M., Ha, J.-W., Kim, S., & Choo, J. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00916 · Cited by 4,163
Collagen Structure and Stability
Shoulders, M. D., & Raines, R. T. (2009) · Annual Review of Biochemistry · doi:10.1146/annurev.biochem.77.032207.120833 · Cited by 3,923
Geometric Deep Learning: Going beyond Euclidean data
Bronstein, M. M., Bruna, J., LeCun, Y., Szlam, A., & Vandergheynst, P. (2017) · IEEE Signal Processing Magazine · doi:10.1109/msp.2017.2693418 · Cited by 3,661 · cited 1×
“Most data used in deep learning can be readily represented in Euclidean space [133], where the convolution operation is properly defined [134].”
BackgroundMoleculeNet: a benchmark for molecular machine learning
Wu, Z., Ramsundar, B., Feinberg, E. N., Gomes, J., Geniesse, C., Pappu, A. S., Leswing, K., & Pande, V. (2018) · Chemical Science · doi:10.1039/c7sc02664a · Cited by 3,107
Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules
Gómez-Bombarelli, R., Wei, J. N., Duvenaud, D., Hernández-Lobato, J. M., Sánchez-Lengeling, B., Sheberla, D., Aguilera-Iparraguirre, J., Hirzel, T. D., Adams, R. P., & Aspuru-Guzik, A. (2018) · ACS Central Science · doi:10.1021/acscentsci.7b00572 · Cited by 3,106
Printing Proteins as Microarrays for High-Throughput Function Determination
MacBeath, G., & Schreiber, S. L. (2000) · Science · doi:10.1126/science.289.5485.1760 · Cited by 2,767
Principles of protein-protein interactions
Jones, S., & Thornton, J. M. (1996) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.93.1.13 · Cited by 2,684 · cited 1×
“Protein–protein interactions compose complexes to conduct numerous biological processes and functions such as metabolic cycles, signal transduction, DNA transcription and replication, catalysis, and immune response [12–18].”
BackgroundConvolutional Networks on Graphs for Learning Molecular Fingerprints
Duvenaud, D., Maclaurin, D., Aguilera-Iparraguirre, J., Gómez-Bombarelli, R., Hirzel, T., Aspuru-Guzik, A., & Adams, R. P. (2015) · arXiv · doi:10.48550/arxiv.1509.09292 · Cited by 2,628
The Database of Interacting Proteins: 2004 update
Salwinski, L. (2004) · Nucleic Acids Research · doi:10.1093/nar/gkh086 · Cited by 2,271
Graph convolutional networks: a comprehensive review
Zhang, S., Tong, H., Xu, J., & Maciejewski, R. (2019) · Computational Social Networks · doi:10.1186/s40649-019-0069-y · Cited by 1,888
Unsupervised Learning of Video Representations using LSTMs
Srivastava, N., Mansimov, E., & Salakhutdinov, R. (2015) · arXiv · doi:10.48550/arxiv.1502.04681 · Cited by 1,861
Oughtred, R., Rust, J., Chang, C., Breitkreutz, B., Stark, C., Willems, A., Boucher, L., Leung, G., Kolas, N., Zhang, F., et al. (2021) · Protein Science · doi:10.1002/pro.3978 · Cited by 1,858 · cited 1×
“There are several known PPI databases, such as Uniprot[106], SWISS-MODEL [107], Negatome 2.0 [108], STRING [109], RCSB PDB [110], BioGRID [111], DIP [112], BIND [113], MINT [114], HPRD [115] and IntAct [116].”
MethodInfluentialGenerating Focused Molecule Libraries for Drug Discovery with Recurrent Neural Networks
Segler, M. H. S., Kogej, T., Tyrchan, C., & Waller, M. P. (2018) · ACS Central Science · doi:10.1021/acscentsci.7b00512 · Cited by 1,537
The coming of age of de novo protein design
Huang, P.-S., Boyken, S. E., & Baker, D. (2016) · Nature · doi:10.1038/nature19946 · Cited by 1,478
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
Orthologs, Paralogs, and Evolutionary Genomics
Koonin, E. V. (2005) · Annual Review of Genetics · doi:10.1146/annurev.genet.39.073003.114725 · Cited by 1,323 · cited 1×
“Orthologs are homologous genes that evolved by vertical descent from a single ancestral gene; In contrast, paralogs evolved by duplication [223,281].”
BackgroundDeep Convolutional Networks on Graph-Structured Data
Henaff, M., Bruna, J., & LeCun, Y. (2015) · arXiv · doi:10.48550/arxiv.1506.05163 · Cited by 1,255
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 1×
“D space [253–255].”
Shapovalov, M. V., & Dunbrack, R. L. (2011) · Structure · doi:10.1016/j.str.2011.03.019 · Cited by 1,000
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 1×
“Building on several recent successful studies using deep learning methods in modelling protein sequences such as contact prediction [359], prediction of secondary structure [360,361], and prediction of the fitness effects of mutations [348], generative modelling methods have begun to show potential…”
BackgroundLu, H., Zhou, Q., He, J., Jiang, Z., Peng, C., Tong, R., & Shi, J. (2020) · Signal Transduction and Targeted Therapy · doi:10.1038/s41392-020-00315-3 · Cited by 891 · cited 2×
“Therefore, identifying protein–protein interactions using deep learning methods helps pave the way towards developing new drugs and targeted therapeutic approaches [8,14,374].”
BackgroundProtein-Protein Interaction Detection: Methods and Analysis
Rao, V. S., Srinivas, K., Sujini, G. N., & Kumar, G. N. S. (2014) · International Journal of Proteomics · doi:10.1155/2014/147648 · Cited by 811
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
Principles for designing ideal protein structures
Koga, N., Tatsumi-Koga, R., Liu, G., Xiao, R., Acton, T. B., Montelione, G. T., & Baker, D. (2012) · Nature · doi:10.1038/nature11600 · Cited by 564
Proof of principle for epitope-focused vaccine design
Correia, B. E., Bates, J. T., Loomis, R. J., Baneyx, G., Carrico, C., Jardine, J. G., Rupert, P., Correnti, C., Kalyuzhniy, O., Vittal, V., et al. (2014) · Nature · doi:10.1038/nature12966 · Cited by 555 · cited 1×
“Some of the successful examples include novel folds in protein design [74,75], enzymes [76,77], antibodies [78–80], vaccines [78,81], ligand-binding proteins [82,83], protein assemblies [84– 88], and membrane proteins [89–91].”
BackgroundSequence 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 1×
“…include, for example, graph neural networks for molecule representation [ 331–334 ], prediction of amino acid sequence for a particular structure using deep neural networks [ 104 ], using GAN to generate DNA sequence [ 335 ], and structure prediction methods using neural networks [ 180,336,337 ].”
MethodGlobal analysis of protein folding using massively parallel design, synthesis, and testing
Rocklin, G. J., Chidyausiku, T. M., Goreshnik, I., Ford, A., Houliston, S., Lemak, A., Carter, L., Ravichandran, R., Mulligan, V. K., Chevalier, A., et al. (2017) · Science · doi:10.1126/science.aan0693 · Cited by 501 · cited 1×
“Even though these methods have shown great achievements, current approaches are unreliable as initial designs frequently fail, entailing multiple trial-and-error cycles [175,176].”
BackgroundVreven, T., Moal, I. H., Vangone, A., Pierce, B. G., Kastritis, P. L., Torchala, M., Chaleil, R., Jiménez-García, B., Bates, P. A., Fernandez-Recio, J., et al. (2015) · Journal of Molecular Biology · doi:10.1016/j.jmb.2015.07.016 · Cited by 424 · cited 1×
“GCN-based [2017] [247] Version 5 of the docking benchmark dataset was used by this study [256], comprising a selected subset of structures generated from X-ray crystallography or nuclear magnetic resonance experiments and containing the atomic coordinates of each amino acid residue in the protein from the Protein Data…”
MethodExpanding functional protein sequence spaces using generative adversarial networks
Repecka, D., Jauniskis, V., Karpus, L., Rembeza, E., Rokaitis, I., Zrimec, J., Poviloniene, S., Laurynenas, A., Viknander, S., Abuajwa, W., et al. (2021) · Nature Machine Intelligence · doi:10.1038/s42256-021-00310-5 · Cited by 388 · cited 1×
“ProteinGAN [352] is a GAN-based method with a customized temporal convolutional network [353] and self-attention mechanism [354] that aims to learn vital long-range inter-residue interactions and sequence motifs, as well as focusing on functional areas [355].”
MethodEnd-to-End Differentiable Learning of Protein Structure
AlQuraishi, M. (2019) · Cell Systems · doi:10.1016/j.cels.2019.03.006 · Cited by 371
Generalized Denoising Auto-Encoders as Generative Models
Bengio, Y., Yao, L., Alain, G., & Vincent, P. (2013) · arXiv · doi:10.48550/arxiv.1305.6663 · Cited by 324
Protein design and variant prediction using autoregressive generative models
Shin, J.-E., Riesselman, A. J., Kollasch, A. W., McMahon, C., Simon, E., Sander, C., Manglik, A., Kruse, A. C., & Marks, D. S. (2021) · Nature Communications · doi:10.1038/s41467-021-22732-w · Cited by 320
Protein-Protein Docking: From Interaction to Interactome
Vakser, I. A. (2014) · Biophysical Journal · doi:10.1016/j.bpj.2014.08.033 · Cited by 318 · cited 2×
“Template-based docking techniques predict PPIs by comparing a protein—protein complex under examination against templates, i.e. other, experimentally determined, protein–protein complex structures [168,211,226–229].”
BackgroundMethodLearning 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
Moal, I. H., & Fernández-Recio, J. (2012) · Bioinformatics · doi:10.1093/bioinformatics/bts489 · Cited by 281
Recurrent Neural Network Model for Constructive Peptide Design
Müller, A. T., Hiss, J. A., & Schneider, G. (2018) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.7b00414 · Cited by 270 · cited 1×
“A LSTM is used in [366] as a generative approach in terms of amino acid sequences and peptide de novo design.”
MethodDeep Learning in Protein Structural Modeling and Design
Gao, W., Mahajan, S. P., Sulam, J., & Gray, J. J. (2020) · Patterns · doi:10.1016/j.patter.2020.100142 · Cited by 227
Porebski, B. T., & Buckle, A. M. (2016) · Protein Engineering Design and Selection · doi:10.1093/protein/gzw015 · Cited by 215
Computational prediction of protein–protein binding affinities
Siebenmorgen, T., & Zacharias, M. (2020) · WIREs Computational Molecular Science · doi:10.1002/wcms.1448 · Cited by 184
Templates are available to model nearly all complexes of structurally characterized proteins
Kundrotas, P. J., Zhu, Z., Janin, J., & Vakser, I. A. (2012) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1200678109 · Cited by 179
Pyrrolysine and Selenocysteine Use Dissimilar Decoding Strategies
Zhang, Y., Baranov, P. V., Atkins, J. F., & Gladyshev, V. N. (2005) · Journal of Biological Chemistry · doi:10.1074/jbc.m501458200 · Cited by 135 · cited 1×
“These are also known as biosynthetic amino acids, and require complex synthetic and translational mechanisms that differ from the canonical enzymatic system used for the 20 standard amino acids [62]) namely, pyrro-lysine [63] and selenocysteine [64].”
BackgroundStrauch, E.-M., Bernard, S. M., La, D., Bohn, A. J., Lee, P. S., Anderson, C. E., Nieusma, T., Holstein, C. A., Garcia, N. K., Hooper, K. A., et al. (2017) · Nature Biotechnology · doi:10.1038/nbt.3907 · Cited by 119 · cited 1×
“Recent advances in computational modelling methods have brought about exceptional findings in protein design, including enzymes [76,77,169], the development of new therapies [170,171], biosensors [172], and small-molecule binders [82].”
BackgroundTargeting protein–protein interactions, a wide open field for drug design
Bakail, M., & Ochsenbein, F. (2016) · Comptes Rendus. Chimie · doi:10.1016/j.crci.2015.12.004 · Cited by 117
Nishimoto, A. T., Sharma, C., & Rogers, P. D. (2020) · Journal of Antimicrobial Chemotherapy · doi:10.1093/jac/dkz400 · Cited by 102
Protein interactions: is seeing believing?
MACKAY, J., SUNDE, M., LOWRY, J., CROSSLEY, M., & MATTHEWS, J. (2007) · Trends in Biochemical Sciences · doi:10.1016/j.tibs.2007.09.006 · Cited by 100
Bittrich, S., Rose, Y., Segura, J., Lowe, R., Westbrook, J. D., Duarte, J. M., & Burley, S. K. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btab813 · Cited by 79
Protein design via deep learning
Ding, W., Nakai, K., & Gong, H. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac102 · Cited by 66
Yugandhar, K., Gupta, S., & Yu, H. (2019) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2019.05.007 · Cited by 51
Zhang, C., Zheng, W., Cheng, M., Omenn, G. S., Freddolino, P. L., & Zhang, Y. (2021) · Journal of Proteome Research · doi:10.1021/acs.jproteome.0c00359 · Cited by 27
Zhang, Y., Bartz, R., Grigoryan, G., Bryant, M., Aaronson, J., Beck, S., Innocent, N., Klein, L., Procopio, W., Tucker, T., et al. (2015) · ACS Chemical Biology · doi:10.1021/cb500759p · Cited by 23
Cited by
Chakraborty, C., Bhattacharya, M., Pal, S., & Lee, S.-S. (2024) · Current Research in Biotechnology · doi:10.1016/j.crbiot.2023.100164 · Cited by 274
Rashwan, A. K., Osman, A. I., Abdelshafy, A. M., Mo, J., & Chen, W. (2025) · Critical Reviews in Food Science and Nutrition · doi:10.1080/10408398.2023.2279696 · Cited by 144
SpatialPPI: Three-dimensional space protein-protein interaction prediction with AlphaFold Multimer
(2024) · doi:10.1016/j.csbj.2024.03.009 · Cited by 29
Seq-InSite: sequence supersedes structure for protein interaction site prediction
Hosseini, S., Golding, G. B., & Ilie, L. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btad738 · Cited by 26
Multimodal graph, surface, and language-based model for protein protein interaction prediction
Arteaga, D., Chervov, N., & Poptsova, M. (2026) · Scientific Reports · doi:10.1038/s41598-025-34758-x · Cited by 3
Metadata
- Authors
- Soleymani, F., Paquet, E., Viktor, H., Michalowski, W., & Spinello, D.
- Venue
- Computational and Structural Biotechnology Journal
- Published
- 2022
- Volume
- 20
- Pages
- 5316-5341
- ISSN
- 2001-0370 (ISSN-L)
- OpenAlex
- W4296293027
- Cited by
- 191
- References
- 64
- FWCI
- 10.63
- Type
- journal article
- Language
- en
- Open access
- gold
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.