The DOI matches its Crossref record; 80 of the 80 references it cites were found in a registry or an index.
Unified rational protein engineering with sequence-based deep representation learning
Open accessPublished version of a preprint
Abstract
Rational protein engineering requires a holistic understanding of protein function. Here, we apply deep learning to unlabeled amino-acid sequences to distill the fundamental features of a protein into a statistical representation that is semantically rich and structurally, evolutionarily and biophysically grounded. We show that the simplest models built on top of this unified representation (UniRep) are broadly applicable and generalize to unseen regions of sequence space. Our data-driven approach predicts the stability of natural and de novo designed proteins, and the quantitative function of molecularly diverse mutants, competitively with the state-of-the-art methods. UniRep further enables two orders of magnitude efficiency improvement in a protein engineering task. UniRep is a versatile summary of fundamental protein features that can be applied across protein engineering informatics.
Overview
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
Cites
Statistics
Verification
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 cited source says what the claim attributes to it.
In the citing paper
“They also learn summary representations, powerful features that can be used to better capture sequence relationships and link sequence to function via transfer learning 27,28,32-35.”
From Learning the protein language: Evolution, structure, and function (2021) · cites this paper as [28]
In the citing paper
“Since then, various language model-based protein embedding methods have been applied to these and other protein prediction problems through transfer learning, including protein phenotype prediction 28,32-34, residue-residue contact prediction 32,70, fold recognition 33, protein-protein 71,72 and protein-drug interaction prediction 35,73.”
From Learning the protein language: Evolution, structure, and function (2021) · cites this paper as [28]
In the citing paper
“In recent years, deep learning architectures that were originally developed for natural language processing (NLP) tasks, such as translating a sentence from one language into another, have been applied successfully to the creation of numerical protein representations from amino acid sequences 17 , 29 .”
From Turnover number predictions for kinetically uncharacterized enzymes using machine and deep learning (2023) · cites this paper as [29]
In the citing paper
“These fingerprints are used together with a 1900-dimensioanl enzyme representation, called UniRep vector 29 , as the input for a gradient boosted decision tree model 34 to predict the K M value for an enzyme-substrate pair.”
From Turnover number predictions for kinetically uncharacterized enzymes using machine and deep learning (2023) · cites this paper as [29]
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
Scikit-learn: Machine Learning in Python
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G., et al. (2012) · arXiv · doi:10.48550/arxiv.1201.0490 · Cited by 71,016 · cited 1×
“We used basic agglomerative clustering (implemented in Python with sklearn 73 ) with the average linkage metric and Euclidean distance for all vector representations, and a precomputed Levenshtein distance matrix using the Python Levenshtein package for the sequence-only control.”
MethodBerman, H. M. (2000) · Nucleic Acids Research · doi:10.1093/nar/28.1.235 · Cited by 40,196 · cited 4×
“We computed the full sequence of hidden states at every position of 3,720 single-domain proteins from the SCOP database 60 (for which we had secondary structure annotation for each of the positions obtained from the PDB 80 ).”
MethodInfluential(2008) · Journal of Machine Learning Research · Cited by 33,051
Kabsch, W., & Sander, C. (1983) · Biopolymers · doi:10.1002/bip.360221211 · Cited by 15,320
On the Properties of Neural Machine Translation: Encoder–Decoder Approaches
Cho, K., van Merriënboer, B., Bahdanau, D., & Bengio, Y. (2014) · SSST@EMNLP · doi:10.3115/v1/w14-4012 · Cited by 7,251
Learning to Forget: Continual Prediction with LSTM
Gers, F. A., Schmidhuber, J., & Cummins, F. (2000) · Neural Computation · doi:10.1162/089976600300015015 · Cited by 6,122
NWChem: A comprehensive and scalable open-source solution for large scale molecular simulations
Valiev, M., Bylaska, E. J., Govind, N., Kowalski, K., Straatsma, T. P., van Dam, H. J. J., Wang, D., Nieplocha, J., Aprà, E., Windus, T. L., et al. (2010) · Computer Physics Communications · doi:10.1016/j.cpc.2010.04.018 · Cited by 5,849
In-Datacenter Performance Analysis of a Tensor Processing Unit
Jouppi, N. P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., Bates, S., Bhatia, S., Boden, N., Borchers, A., et al. (2017) · Proceedings of the 44th Annual International Symposium on Computer Architecture · doi:10.1145/3079856.3080246 · Cited by 4,537
An Overview of Multi-Task Learning in Deep Neural NetworksPreprint
Ruder, S. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1706.05098 · Cited by 3,827
Molecular dynamics simulations of biomolecules
Karplus, M., & McCammon, J. (2002) · Nature Structural Biology · doi:10.1038/nsb0902-646 · Cited by 3,401
Research works printed but not linked: 8
- Brookes, D. H., Park, H. & Listgarten, J. Conditioning by adaptive sampling for robust design. Proc. Machine Learn. Res. 97, 773–782 (2019).
- Hernández-Lobato, J. M., Requeima, J., Pyzer-Knapp, E. O. & Aspuru-Guzik, A. Parallel and distributed thompson sampling for large-scale accelerated exploration of chemical space. Preprint at arXiv https://arxiv.org/abs/1706.01825 (2017). · arXiv
- Snoek, J., Larochelle, H. & Adams, R. P. in Advances in Neural Information Processing Systems Vol. 25 (eds. Pereira, F. et al.) 2951–2959 (Curran Associates, Inc., 2012).
- Griffiths, R.-R. & Hernández-Lobato, J. M. Constrained Bayesian optimization for automaticchemical design. Preprint at arXiv https://arxiv.org/abs/1709.05501 (2017). · arXiv
- Yang, K. K., Chen, Y., Lee, A. & Yue, Y. Batched stochastic Bayesian optimization via combinatorial constraints design. Preprint at arXiv https://arxiv.org/abs/1904.08102 (2019). · arXiv
- González, J., Longworth, J., James, D. C. & Lawrence, N. D. Bayesian optimization for synthetic gene design. Preprint at arXiv https://arxiv.org/abs/1505.01627 (2015). · arXiv
- Jouppi, N. P. et al. In-datacenter performance analysis of a tensorprocessing unit. In Proc. 44th Annual International Symposium of Computer Architecture Vol. 45, 1–12 (ACM, 2017).
- Salimans, T. & Kingma, D. P. Weight normalization: a simple reparameterization to accelerate training of deep neural networks. Preprint at arXiv https://arxiv.org/abs/1602.07868 (2016). · arXiv
Cited by
Sorted by citations
Items 201–300 of 1,003
SYNBIP: synthetic binding proteins for research, diagnosis and therapy
Wang, X., Li, F., Qiu, W., Xu, B., Li, Y., Lian, X., Yu, H., Zhang, Z., Wang, J., Li, Z., et al. (2021) · Nucleic Acids Research · doi:10.1093/nar/gkab926 · Cited by 70
Theoretical and Data-Driven Approaches for Biomolecular Condensates
Saar, K. L., Qian, D., Good, L. L., Morgunov, A. S., Collepardo‐Guevara, R., Best, R. B., & Knowles, T. P. J. (2023) · Chemical Reviews · doi:10.1021/acs.chemrev.2c00586 · Cited by 66
Protein language models are biased by unequal sequence sampling across the tree of life
Ding, F., & Steinhardt, J. (2024) · doi:10.1101/2024.03.07.584001 · Cited by 57
“…al., 2022; Ferruz et al., 2022; Elnaggar et al., 2023; Lin et al., 2023; Nijkamp et al., 2023), CNNs (Yang et al., 2022), and other architectures (Alley et al., 2019). pLMs can generate sequences that successfully fold into functional proteins, both by sampling from the model unconditionally,…”
BackgroundMoffat, L., & Jones, D. T. (2021) · Bioinformatics · doi:10.1093/bioinformatics/btab491 · Cited by 73
Transformer-based protein generation with regularized latent space optimization
Castro, E., Godavarthi, A., Rubinfien, J., Givechian, K. B., Bhaskar, D., & Krishnaswamy, S. (2022) · Nature Machine Intelligence · doi:10.1038/s42256-022-00532-1 · Cited by 76
“While other encoding methods that rely on convolutional and recurrent architectures have demonstrated success in this domain 14,15 , we chose to use a transformer for several key reasons.”
Protein engineering via Bayesian optimization-guided evolutionary algorithm and robotic experiments
Hu, R., Fu, L., Chen, Y., Chen, J., Qiao, Y., & Si, T. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac570 · Cited by 71
Machine learning prediction of enzyme optimum pH
Gado, J. E., Knotts, M., Shaw, A. Y., Marks, D. S., Gauthier, N. P., Sander, C., & Beckham, G. T. (2025) · Nature Machine Intelligence · doi:10.1038/s42256-025-01026-6 · Cited by 78
“While truncated backpropagation through time may help to alleviate vanishing gradients 96 , evaluation of RNNs with truncated backpropagation through time was beyond the scope of this work.”
Heterogeneity of the GFP fitness landscape and data-driven protein design
Somermeyer, L. G., Fleiss, A., Mishin, A. S., Bozhanova, N. G., Igolkina, A. A., Meiler, J., Pujol, M.-E. A., Putintseva, E. V., Sarkisyan, K. S., & Kondrashov, F. A. (2022) · eLife · doi:10.7554/elife.75842 · Cited by 75
Hoarfrost, A., Aptekmann, A. A., Farfañuk, G., & Bromberg, Y. (2022) · Nature Communications · doi:10.1038/s41467-022-30070-8 · Cited by 65
Predicting the antigenic evolution of SARS-COV-2 with deep learning
Han, W., Chen, N., Xu, X., Sahil, A., Zhou, J., Li, Z., Zhong, H., Gao, E., Zhang, R., Wang, Y., et al. (2023) · Nature Communications · doi:10.1038/s41467-023-39199-6 · Cited by 66
Interpretable detection of novel human viruses from genome sequencing data
Bartoszewicz, J. M., Seidel, A., & Renard, B. Y. (2021) · NAR Genomics and Bioinformatics · doi:10.1093/nargab/lqab004 · Cited by 67
Improving Enzyme Optimum Temperature Prediction with Resampling Strategies and Ensemble Learning
Gado, J. E., Beckham, G. T., & Payne, C. M. (2020) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.0c00489 · Cited by 64
Cluster learning-assisted directed evolution
Qiu, Y., Hu, J., & Wei, G. (2021) · Nature Computational Science · doi:10.1038/s43588-021-00168-y · Cited by 63
“Adapted from natural language process55 ing, many model architectures, such as variational auto-encoder [15], recurrent neural network [16, 17], and 56 transformer [18], can be used to train the protein embedding models [13].”
MethodDeep generative design of RNA family sequences
Sumi, S., Hamada, M., & Saito, H. (2024) · Nature Methods · doi:10.1038/s41592-023-02148-8 · Cited by 67
“To examine the effects in generative capacity by explicitly using a CM, we first compared RfamGen to three ablation models that do not take by a sequence representation 19–22 .”
Bachas, S. T., Rakocevic, G., Spencer, D., Sastry, A. V., Haile, R. G., Sutton, J. M., Kasun, G. W., Stachyra, A. L., Gutierrez, J. M., Yassine, E., et al. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.08.16.504181 · Cited by 63
Identification of sub-Golgi protein localization by use of deep representation learning features
Lv, Z., Wang, P., Zou, Q., & Jiang, Q. (2020) · Bioinformatics · doi:10.1093/bioinformatics/btaa1074 · Cited by 60
Persistent spectral theory-guided protein engineering
(2023) · doi:10.1038/s43588-022-00394-y · Cited by 62
“Two existing LSTM-based models, UniRep [14] and Bepler [13], were also implemented.”
BackgroundMethodInfluentialRevolutionizing enzyme engineering through artificial intelligence and machine learning
Singh, N., Malik, S., Gupta, A., & Srivastava, K. R. (2021) · Emerging Topics in Life Sciences · doi:10.1042/etls20200257 · Cited by 60
Lee, J., Jun, D. W., Song, I., & Kim, Y. (2024) · Journal of Cheminformatics · doi:10.1186/s13321-024-00808-1 · Cited by 62
Deep learning frameworks for protein–protein interaction prediction
Hu, X., Feng, C., Ling, T., & Chen, M. (2022) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2022.06.025 · Cited by 61
“[57] uses the UniRep model [58] to learn the representations of the corresponding proteins.”
MethodAccelerating drug target inhibitor discovery with a deep generative foundation model
Chenthamarakshan, V., Hoffman, S. C., Owen, C. D., Lukacik, P., Strain-Damerell, C. M., Fearon, D., Malla, T. R., Tumber, A., Schofield, C. J., Duyvesteyn, H. M. E., et al. (2023) · Science Advances · doi:10.1126/sciadv.adg7865 · Cited by 65
“Further, a target-chemical binder (strong/weak) predictor was trained on the latent z vectors of chemicals and the pretrained protein sequence embeddings (15), which used the data released as part of DeepAffinity (59).”
MethodEvaluating generalizability of artificial intelligence models for molecular datasets
Ektefaie, Y., Shen, A. N., Bykova, D., Marin, M. G., Žitnik, M., & Farhat, M. R. (2024) · Nature Machine Intelligence · doi:10.1038/s42256-024-00931-6 · Cited by 69
Genetically Encodable Fluorescent and Bioluminescent Biosensors Light Up Signaling Networks
Zhou, X., Mehta, S., & Zhang, J. (2020) · Trends in Biochemical Sciences · doi:10.1016/j.tibs.2020.06.001 · Cited by 57
Improved Bst DNA Polymerase Variants Derived via a Machine Learning Approach
Paik, I., Ngo, P. H. T., Shroff, R., Diaz, D. J., Maranhao, A. C., Walker, D. J. F., Bhadra, S., & Ellington, A. D. (2021) · Biochemistry · doi:10.1021/acs.biochem.1c00451 · Cited by 57
Wittmund, M., Cadet, F., & Davari, M. D. (2022) · ACS Catalysis · doi:10.1021/acscatal.2c01426 · Cited by 62
In vitro continuous protein evolution empowered by machine learning and automation
Yu, T., Boob, A. G., Singh, N., Su, Y., & Zhao, H. (2023) · Cell Systems · doi:10.1016/j.cels.2023.04.006 · Cited by 55
Generative models for protein sequence modeling: recent advances and future directions
Mardikoraem, M., Wang, Z., Pascual, N., & Woldring, D. (2023) · Briefings in Bioinformatics · doi:10.1093/bib/bbad358 · Cited by 61
“TAPE[25],UniRep[26],ESM [27] and ProtTrans [28] are among the many successful studies that have applied NLP techniques to learn the dependencies in protein sequences and be able to numerically represent them in fixed vector formats (i.e. embeddings) that are rich in semantics and syntax information.”
BackgroundNext generation of multispecific antibody engineering
Kéri, D., Walker, M., Singh, I., Nishikawa, K. K., & Garcés, F. (2023) · Antibody Therapeutics · doi:10.1093/abt/tbad027 · Cited by 59
Notin, P., van Niekerk, L., Kollasch, A. W., Ritter, D., Gal, Y., & Marks, D. S. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.12.07.519495 · Cited by 62
“…had to be used for predictions, giving rise to ‘hybrid models’ that borrow strengths from family-specific and family-agnostic approaches: Unirep [Alley et al., 2019] and ESM-1v [Meier et al., 2021] rely on fine-tuning pretrained models on an MSA for the protein of interest; the MSA…”
BackgroundMethodInfluentialLeander, M., Liu, Z., Cui, Q., & Raman, S. (2022) · eLife · doi:10.7554/elife.79932 · Cited by 54
Chen, X., Li, C., Bernards, M. T., Shi, Y., Shao, Q., & He, Y. (2021) · Molecular Systems Design & Engineering · doi:10.1039/d0me00161a · Cited by 56
Rembeza, E., & Engqvist, M. K. M. (2021) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1009446 · Cited by 55
“The sequence interrelatedness of these diverse proteins was visualized in a multidimensional scaling (MDS) plot using computed UniRep embeddings [37]; a smaller distance in this plot indicates higher relatedness (Figs 1 and S2).”
MethodComputational enzyme redesign: large jumps in function
Cui, Y., Sun, J., & Wu, B. (2022) · Trends in Chemistry · doi:10.1016/j.trechm.2022.03.001 · Cited by 53
Ao, Y., Pei, S., Xiang, C., Menke, M. J., Shen, L. F., Sun, C., Dörr, M., Born, S., Höhne, M., & Bornscheuer, U. T. (2023) · Angewandte Chemie International Edition · doi:10.1002/anie.202301660 · Cited by 54
Dewaker, V., Morya, V. K., Kim, Y. H., Park, S. T., Kim, H. S., & Koh, Y. H. (2025) · Biomarker Research · doi:10.1186/s40364-025-00764-4 · Cited by 62
Generative power of a protein language model trained on multiple sequence alignments
(2023) · doi:10.7554/elife.79854 · Cited by 43
“The great success of supervised protein structure prediction by AlphaFold [28] is partly based on the use of transformers.”
Proteochemometrics – recent developments in bioactivity and selectivity modeling
Bongers, B. J., IJzerman, A. P., & Van Westen, G. J. P. (2019) · Drug Discovery Today: Technologies · doi:10.1016/j.ddtec.2020.08.003 · Cited by 54
The generative capacity of probabilistic protein sequence models
McGee, F. A., Hauri, S., Novinger, Q., Vučetić, S., Levy, R. M., Carnevale, V., & Haldane, A. (2021) · Nature Communications · doi:10.1038/s41467-021-26529-9 · Cited by 53
Linguistically inspired roadmap for building biologically reliable protein language models
Vu, M. H., Akbar, R., Robert, P. A., Świątczak, B., Sandve, G. K., Greiff, V., & Haug, D. T. T. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-023-00637-1 · Cited by 59
“All studies clustered token embeddings in terms of physicochemical properties 6,9,58,79 , as they are well-defined, and none examined whether the tokens can capture more abstract functional meaning, which would have been illuminating.”
BackgroundMethodInfluentialA Perspective on Protein Structure Prediction Using Quantum Computers
Doğa, H., Raubenolt, B. A., Cumbo, F., Joshi, J., DiFilippo, F. P., Qin, J., Blankenberg, D. J., & Shehab, O. (2024) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.4c00067 · Cited by 67
Cross-disciplinary perspectives on the potential for artificial intelligence across chemistry
Mroz, A. M., Basford, A. R., Hastedt, F., Jayasekera, I. S., Mosquera‐Lois, I., Sedgwick, R., Ballester, P. J., Bocarsly, J. D., del Rio‐Chanona, E. A., Evans, M. L., et al. (2025) · Chemical Society Reviews · doi:10.1039/d5cs00146c · Cited by 61
Large language models for drug discovery and development
Zheng, Y., Koh, H. Y., Ju, J., Yang, M., May, L. T., Webb, G. I., Li, L., Pan, S., & Church, G. M. (2025) · Patterns · doi:10.1016/j.patter.2025.101346 · Cited by 121
“54 Earlier work, such as UniRep, 55 demonstrated how deep sequence-based representation learning enables efficient exploration of evolutionary fitness landscapes, improving stability and functional annotation predictions.”
Self-citationProT-VAE: Protein Transformer Variational AutoEncoder for functional protein design
Sevgen, E., Moller, J., Lange, A., Parker, J., Quigley, S. D., Mayer, J., Srivastava, P., Gayatri, S., Hosfield, D. J., Dilks, C. M., et al. (2025) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2408737122 · Cited by 59
Pre-Training of Deep Bidirectional Protein Sequence Representations With Structural Information
Min, S., Park, S., Kim, S., Choi, H.-S., Lee, B., & Yoon, S. (2021) · IEEE Access · doi:10.1109/access.2021.3110269 · Cited by 59
SETH predicts nuances of residue disorder from protein embeddings
Ilzhöfer, D., Heinzinger, M., & Rost, B. (2022) · Frontiers in Bioinformatics · doi:10.3389/fbinf.2022.1019597 · Cited by 52
Shi, Z., Deng, R., Yuan, Q., Mao, Z., Wang, R., Li, H., Liao, X., & Ma, H. (2023) · Research · doi:10.34133/research.0153 · Cited by 53
“2B, our method can achieve scores of 92.01% and 93.61% in terms of F1 score in the above-constructed testing sets, respectively.”
InfluentialBeyond natural: synthetic expansions of botanical form and function
Patron, N. J. (2020) · New Phytologist · doi:10.1111/nph.16562 · Cited by 53
Using machine learning to predict the effects and consequences of mutations in proteins
Diaz, D. J., Kulikova, A. V., Ellington, A. D., & Wilke, C. O. (2023) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2022.102518 · Cited by 54
Serverless Prediction of Peptide Properties with Recurrent Neural Networks
Ansari, M., & White, A. D. (2023) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.2c01317 · Cited by 54
UMI-linked consensus sequencing enables phylogenetic analysis of directed evolution
Zurek, P. J., Knyphausen, P., Neufeld, K., Pushpanath, A., & Hollfelder, F. (2020) · Nature Communications · doi:10.1038/s41467-020-19687-9 · Cited by 47
Learning protein fitness landscapes with deep mutational scanning data from multiple sources
Chen, L., Zhang, Z., Li, Z., Li, R., Huo, R., Chen, L., Wang, D., Luo, X., Chen, K., Liao, C., et al. (2023) · Cell Systems · doi:10.1016/j.cels.2023.07.003 · Cited by 57
Protein subcellular localization prediction tools
Gillani, M., & Pollastri, G. (2024) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2024.04.032 · Cited by 52
Feature Reuse and Scaling: Understanding Transfer Learning with Protein Language Models
Li, F.-Z., Amini, A. P., Yue, Y., Yang, K. K., & Lu, A. X. (2024) · doi:10.1101/2024.02.05.578959 · Cited by 51
“…pretrained PLMs have been proposed in the past few years (Rives et al., 2021; Yang et al., 2022; El-naggar et al., 2022; Brandes et al., 2022; Alley et al., 2019; Rao et al., 2019; Elnaggar et al., 2023; Lin et al., 2023), these works primarily focus on validating that pretraining improves…”
BackgroundAn analysis of protein language model embeddings for fold prediction
Villegas-Morcillo, A., Gomez, A. M., & Sanchez, V. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac142 · Cited by 46
Prediction of designer-recombinases for DNA editing with generative deep learning
Schmitt, L. T., Paszkowski-Rogacz, M., Jug, F., & Buchholz, F. (2022) · Nature Communications · doi:10.1038/s41467-022-35614-6 · Cited by 45
“Beyond mere generation of new protein sequences, the goal is often to generate variants with improved properties (Alley et al. 2019; Linder et al. 2020; Biswas et al. 2021; Hawkins-Hooker et al. 2021; Saka et al. 2021; Shin et al. 2021).”
BackgroundMethodAdvancing Targeted Protein Degradation via Multiomics Profiling and Artificial Intelligence
Duran-Frigola, M., Cigler, M., & Winter, G. E. (2023) · Journal of the American Chemical Society · doi:10.1021/jacs.2c11098 · Cited by 50
Yang, M., Huang, L., Huang, H., Tang, H., Zhang, N., Yang, H., Wu, J., & Mu, F. (2022) · Nucleic Acids Research · doi:10.1093/nar/gkac326 · Cited by 56
The RESP AI model accelerates the identification of tight-binding antibodies
Parkinson, J., Hard, R. L., & Wang, W. (2023) · Nature Communications · doi:10.1038/s41467-023-36028-8 · Cited by 52
“To evaluate this more critical metric, we encode the WT and mutant Atezolizumab library (construction described below) using the auto-encoder described above, one-hot encoding, the ProtVec encoding scheme, the UniRep encoding scheme, the FAIR-ESM encoding scheme, the AbLang seq-coding scheme and the AntiBertY embed-di…”
MethodInfluentialAddressing epistasis in the design of protein function
Lipsh‐Sokolik, R., & Fleishman, S. J. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2314999121 · Cited by 55
Livesey, B. J., & Marsh, J. A. (2025) · Genome Biology · doi:10.1186/s13059-025-03575-w · Cited by 50
A trimodal protein language model enables advanced protein searches
Su, J., He, Y., You, S., Jiang, S. J., Zhou, X., Zhang, X., Wang, Y., Su, X., Tolstoy, I., Chang, X., et al. (2025) · Nature Biotechnology · doi:10.1038/s41587-025-02836-0 · Cited by 67
Lu, A. X., Lu, A. X., Pritišanac, I., Zarin, T., Forman-Kay, J. D., & Moses, A. M. (2022) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1010238 · Cited by 56
Duong‐Trung, N., Born, S., Kim, J. W., Schermeyer, M., Paulick, K., Borisyak, M., Bournazou, M. N. C., Werner, T., Scholz, R., Schmidt-Thieme, L., et al. (2022) · Biochemical Engineering Journal · doi:10.1016/j.bej.2022.108764 · Cited by 56
Accelerating therapeutic protein design with computational approaches toward the clinical stage
Chen, Z., Wang, X., Chen, X., Huang, J., Wang, C., Wang, J., & Wang, Z. (2023) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2023.04.027 · Cited by 49
“In the research of Ethan C. Alley and co-workers, they carry out a pattern of unsupervised learning, next token prediction, with a recurrent neural network trained by the UniRef50 dataset with 24 million amino acid sequences [104].”
Harnessing the power of artificial intelligence to advance cell therapy
Capponi, S., & Daniels, K. G. (2023) · Immunological Reviews · doi:10.1111/imr.13236 · Cited by 51
Neural network extrapolation to distant regions of the protein fitness landscape
Freschlin, C. R., Fahlberg, S. A., Heinzelman, P., & Romero, P. A. (2024) · Nature Communications · doi:10.1038/s41467-024-50712-3 · Cited by 52
DLPacker : Deep learning for prediction of amino acid side chain conformations in proteins
Misiura, M., Shroff, R., Thyer, R., & Kolomeisky, A. B. (2022) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.26311 · Cited by 51
pLM4Alg: Protein Language Model-Based Predictors for Allergenic Proteins and Peptides
Du, Z., Xu, Y., Liu, C., & Li, Y. (2023) · Journal of Agricultural and Food Chemistry · doi:10.1021/acs.jafc.3c07143 · Cited by 48
Are protein language models the new universal key?
Weißenow, K., & Rost, B. (2025) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2025.102997 · Cited by 54
Geometricus represents protein structures as shape-mers derived from moment invariants
Durairaj, J., Akdel, M., de Ridder, D., & van Dijk, A. D. J. (2020) · Bioinformatics · doi:10.1093/bioinformatics/btaa839 · Cited by 47
Recent Advances in Machine Learning Variant Effect Prediction Tools for Protein Engineering
Horne, J., & Shukla, D. (2022) · Industrial & Engineering Chemistry Research · doi:10.1021/acs.iecr.1c04943 · Cited by 48
Harmalkar, A., Rao, R., Xie, Y. R., Honer, J., Deisting, W., Anlahr, J., Hoenig, A., Czwikla, J., Sienz-Widmann, E., Rau, D., et al. (2023) · mAbs · doi:10.1080/19420862.2022.2163584 · Cited by 48
“Leveraging sequence and structural information as features, ML approaches applied on a plethora of prediction tasks, e.g., fluorescence landscapes, 16–18 intrinsic stability, 19,20 missense variant effects, 21 protein fitness, 22 antigen-specificity, 23 have shown high success rates.”
MethodExperimental characterization of de novo proteins and their unevolved random-sequence counterparts
Heames, B., Buchel, F., Aubel, M., Tretyachenko, V., Loginov, D. S., Novák, P., Lange, A., Bornberg‐Bauer, E., & Hlouchová, K. (2023) · Nature Ecology & Evolution · doi:10.1038/s41559-023-02010-2 · Cited by 46
“To predict stability scores, we used an implementation of UniRep(42,60) to generate sequence embeddings of size 1,900 and trained a sparse linear model (Lasso least-angle regression with tenfold cross-validation) on a dataset of de novo-designed proteins with experimentally determined stability scores(41), as describe…”
MethodAbdelkader, G. A., & Kim, J. (2024) · Current Drug Targets · doi:10.2174/0113894501330963240905083020 · Cited by 49
Fenoy, E., Edera, A. A., & Stegmayer, G. S. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac232 · Cited by 47
Manfredi, M., Savojardo, C., Martelli, P. L., & Casadio, R. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac678 · Cited by 48
How to approach machine learning-based prediction of drug/compound–target interactions
Ataş, H., & Doğan, T. (2023) · Journal of Cheminformatics · doi:10.1186/s13321-023-00689-w · Cited by 48
“Even though they are trained solely on the information about the arrangement of amino acids in the sequence, these models are still found to be successful in automatically extracting physicochemical properties [23] and functional characteristics of proteins [24].”
BackgroundData‐Driven Protein Engineering for Improving Catalytic Activity and Selectivity
Ao, Y., Dörr, M., Menke, M. J., Born, S., Heuson, E., & Bornscheuer, U. T. (2023) · ChemBioChem · doi:10.1002/cbic.202300754 · Cited by 44
ProLLaMA: A Protein Large Language Model for Multitask Protein Language Processing
Lv, L., Lin, Z., Li, H., Liu, Y., Cui, J., Chen, C. Y., Yuan, L., & Tian, Y. (2025) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2025.3564914 · Cited by 93
Protein Design with Deep Learning
Defresne, M., Barbe, S., & Schiex, T. (2021) · International Journal of Molecular Sciences · doi:10.3390/ijms222111741 · Cited by 48
“[37] who found a better performance on buried residues [56,57], which is consistent with the fact that these residues are more constrained.”
A language model beats alphafold2 on orphans
Michaud, J. M., Madani, A., & Fraser, J. S. (2022) · Nature Biotechnology · doi:10.1038/s41587-022-01466-0 · Cited by 45
Drug-target interaction/affinity prediction: Deep learning models and advances review
Vefghi, A., Rahmati, Z., & Akbari, M. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.110438 · Cited by 47
Geffen, Y., Ofran, Y., & Unger, R. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac474 · Cited by 40
Fast, sensitive detection of protein homologs using deep dense retrieval
Hong, L., Hu, Z., Sun, S., Tang, X., Wang, J., Tan, Q., Zheng, L., Wang, S., Xu, S. Y., King, I., et al. (2024) · Nature Biotechnology · doi:10.1038/s41587-024-02353-6 · Cited by 46
Data-driven computational protein design
Frappier, V., & Keating, A. E. (2021) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2021.03.009 · Cited by 44
Mohapatra, S., An, J., & Gómez‐Bombarelli, R. (2022) · Machine Learning Science and Technology · doi:10.1088/2632-2153/ac545e · Cited by 47
“For instance, unsupervised learning of protein sequences has resulted in functional annotation and identification of sub-families for yet unseen sequences [37], in addition to creating information-rich embeddings for downstream data-poorer property prediction tasks [36, 38].”
BackgroundDigitally enabled approaches for the scale up of mammalian cell bioreactors
Alavijeh, M. K., Baker, I., Lee, Y. Y., & Gras, S. L. (2022) · Digital Chemical Engineering · doi:10.1016/j.dche.2022.100040 · Cited by 44
Learning deep representations of enzyme thermal adaptation
Li, G., Buric, F., Zrimec, J., Viknander, S., Nielsen, J., Zelezniak, A., & Engqvist, M. K. M. (2022) · Protein Science · doi:10.1002/pro.4480 · Cited by 44
“In Figure 3a and 3b, the R 2 score of three best regression models were shown for iFeature and UniRep.”
MethodNearest neighbor search on embeddings rapidly identifies distant protein relations
Schütze, K., Heinzinger, M., Steinegger, M., & Rost, B. (2022) · Frontiers in Bioinformatics · doi:10.3389/fbinf.2022.1033775 · Cited by 41
“While the idea for this is not new (Alley et al., 2019; Littmann et al., 2021a; Littmann et al., 2021b; Ofer et al., 2021; Bileschi et al., 2022; Heinzinger et al., 2022; Nallapareddy et al., 2022), here we presented a more rigorous and generic framework for directly comparing embeddingbased to sequence-based alignmen…”
BackgroundInfluentialProGen: Language Modeling for Protein GenerationPreprint
Madani, A., McCann, B., Naik, N., Keskar, N. S., Anand, N., Eguchi, R. R., Huang, P., & Socher, R. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.03497 · Cited by 42
Deep neural language modeling enables functional protein generation across familiesPreprint
Madani, A., Krause, B., Greene, E. R., Subramanian, S., Mohr, B. P., Holton, J. M., Olmos, J. L., Xiong, C., Sun, Z. Z., Socher, R., et al. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.07.18.452833 · Cited by 47
“In contrast to transformer-based language models that encode amino acid sequences for discriminative protein prediction tasks 25,33,34 , ProGen is a decoder transformer tailored for autoregressive generation.”
ResultTherapeutic enzyme engineering using a generative neural network
Giessel, A. J., Dousis, A. D., Ravichandran, K. R., Smith, K., Sur, S., McFadyen, I. J., Zheng, W., & Licht, S. (2022) · Scientific Reports · doi:10.1038/s41598-022-05195-x · Cited by 41
Upadhyay, R., Phlypo, R., Saini, R., & Liwicki, M. (2024) · IEEE Access · doi:10.1109/access.2024.3478805 · Cited by 53
Cui, F., Zhang, Z., Cao, C., Zou, Q., Chen, D., & Su, X. (2022) · PROTEOMICS · doi:10.1002/pmic.202100197 · Cited by 40
100th Anniversary of Macromolecular Science Viewpoint: Data-Driven Protein Design
Ferguson, A. L., & Ranganathan, R. (2021) · ACS Macro Letters · doi:10.1021/acsmacrolett.0c00885 · Cited by 40
Using AlphaFold for Rapid and Accurate Fixed Backbone Protein Design
Moffat, L., Greener, J. G., & Jones, D. T. (2021) · doi:10.1101/2021.08.24.457549 · Cited by 33
“…variational autoencoders (Greener et al., 2018; Hawkins-Hooker et al., 2021), deep exploration networks (Linder et al., 2020), graph neural networks (Strokach et al., 2020), recurrent neural networks (Alley et al., 2019) and autoregressive models (Shin et al., 2021; Trinquier et al., 2021).”
BackgroundApplications of Machine and Deep Learning in Adaptive Immunity
Pertseva, M., Gao, B., Neumeier, D., Yermanos, A., & Reddy, S. T. (2021) · Annual Review of Chemical and Biomolecular Engineering · doi:10.1146/annurev-chembioeng-101420-125021 · Cited by 38
Aparicio, J. A. B., Olivares‐Gil, A., Rodríguez, J. J., García‐Osorio, C., & Díez-Pastor, J.-F. (2023) · Information Fusion · doi:10.1016/j.inffus.2023.102035 · Cited by 41
Menke, M. J., Ao, Y., & Bornscheuer, U. T. (2024) · ACS Catalysis · doi:10.1021/acscatal.4c00987 · Cited by 38
Xu, Y., Liu, D., & Gong, H. (2024) · Nature Computational Science · doi:10.1038/s43588-024-00716-2 · Cited by 41
Metadata
- Authors
- Alley, E. C., Khimulya, G., Biswas, S., AlQuraishi, M., & Church, G. M.
- Affiliations
- Harvard University (US)
- Corresponding author
- Church, George M.
- Venue
- Nature Methods
- Published
- 2019-10-21
- Volume
- 16
- Issue
- 12
- Pages
- 1315-1322
- ISSN
- 1548-7091 (ISSN-L)
- 1548-7105
- PMC
- PMC7067682
- OpenAlex
- W2980789587
- Cited by
- 1,131
- References
- 80
- Citations by year
- 2026: 130 · 2025: 170 · 2024: 201 · 2023: 197 · 2022: 188 · 2021: 159 · 2020: 69 · 2019: 17
- FWCI
- 35.71
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Licence
- Locations
- Nature Methods (publishedVersion)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
- Research Square (acceptedVersion, cc-by)
- PubMed (publishedVersion)
- Europe PMC (PubMed Central) (acceptedVersion)
- PubMed Central (acceptedVersion)
- OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
- Funders
- Amazon Web Services Cloud Credits for Research
- Center for Effective Altruism
- U.S. Department of Health & Human Services | National Institutes of Health (T32HG002295, U54-CA225088)
- National Science Foundation GRFP Fellowship
- U.S. Department of Health & Human Services | NIH | National Institute of General Medical Sciences (P50GM107618)
- 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
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › RNA and protein synthesis mechanisms
- Keywords
- rational protein design, protein stability prediction, de novo protein design, protein mutants, recurrent neural networks, protein function prediction
- MeSH terms
- Deep Learning, Amino Acid Sequence, Mutation, Protein Engineering, Protein Stability
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
