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 501–600 of 1,003
Evotuning protocols for Transformer-based variant effect prediction on multi-domain proteins
Yamaguchi, H., & Saitō, Y. (2021) · Briefings in Bioinformatics · doi:10.1093/bib/bbab234 · Cited by 15
Jiang, Z., Xiao, S.-R., & Liu, R. (2021) · Briefings in Bioinformatics · doi:10.1093/bib/bbab411 · Cited by 15
“Recently, deep learning frameworks have been commonly applied to the recognition of protein functional residues [78, 79] and the prediction of structural properties for RNA [38, 74], so these methodologies could be used to predict binding nucleotides in NAs when more structures become available.”
MethodDeep Learning-Powered Prediction of Human-Virus Protein-Protein Interactions
Yang, X., Yang, S., Ren, P., Wuchty, S., & Zhang, Z. (2022) · Frontiers in Microbiology · doi:10.3389/fmicb.2022.842976 · Cited by 16
“Using a pre-trained UNIREP model (Alley et al., 2019) based on multiplicative LSTM (mLSTM) human and viral protein embeddings were obtained to predict human-specific and human-virus PPIs based on known PPI data from various benchmark datasets.”
MethodMaeda, K., Hatae, A., Sakai, Y., Boogerd, F. C., & Kurata, H. (2022) · BMC Bioinformatics · doi:10.1186/s12859-022-05009-x · Cited by 19
“In their approach, sub-strate’s structure and enzyme’s amino acid sequence were converted into a 52-dimen-sional fingerprint vector and a 1,900-dimensional UniRep [35] vector, respectively.”
MethodProtein Function Prediction With Functional and Topological Knowledge of Gene Ontology
Zhao, Y., Yang, Z., Hong, Y., Yang, Y., Wang, L., Zhang, Y., Lin, H., & Wang, J. (2023) · IEEE Transactions on NanoBioscience · doi:10.1109/tnb.2023.3278033 · Cited by 15
Wang, X. R., Zhou, W., Wang, R., & Gao, X. (2024) · Symmetry · doi:10.3390/sym16040464 · Cited by 17
Enhancing cryo-EM structure prediction with DeepTracer and AlphaFold2 integration
Chen, J., Zia, A., Luo, A., Meng, H., Wang, F., Hou, J., Cao, R., & Si, D. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae118 · Cited by 19
Machine learning for predicting protein properties: A comprehensive review
Wang, Y., Zhang, Y., Zhan, X., He, Y., Yang, Y., Cheng, L., & Alghazzawi, D. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128103 · Cited by 15
Capturing Protein Domain Structure and Function Using Self-Supervision on Domain Architectures
Melidis, D. P., Malone, B., & Nejdl, W. (2021) · Algorithms · doi:10.3390/a14010028 · Cited by 15
Wang, Q., Wang, B., Xu, Z., Wu, J., Zhao, P., Li, Z., Wang, S., Huang, J., & Cui, S. R. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i1.16141 · Cited by 18
“Recent work in protein sequence pre-training has shown positive results on various downstream tasks including secondary structure prediction (Alley et al. 2019; Bepler and Berger 2019; Heinzinger et al. 2019; Rao et al. 2019; Rives et al. 2019).”
BackgroundEnergetic Basis and Design of Enzyme Function Demonstrated Using GFP, an Excited-State Enzyme
Lin, C., Romei, M. G., Mathews, I. I., & Boxer, S. G. (2022) · Journal of the American Chemical Society · doi:10.1021/jacs.1c12305 · Cited by 17
Melo, M. C. R., & Bernardi, R. C. (2023) · Biophysical Journal · doi:10.1016/j.bpj.2023.01.042 · Cited by 15
A primer on model-guided exploration of fitness landscapes for biological sequence designPreprint
Sinai, S., & Kelsic, E. D. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.10614 · Cited by 29
Generating novel protein sequences using Gibbs sampling of masked language modelsPreprint
Johnson, S. R., Monaco, S., Massie, K., & Syed, Z. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.01.26.428322 · Cited by 18
“…is an RNN model trained on a subset of UniProt; one key finding from the UniRep work is that using the pre-training task (in their case, next token prediction) to fine-tune on additional examples of proteins from a particular family, increases performance on down-stream tasks for that family [1, 5].”
MethodImproving Generalizability of Protein Sequence Models with Data AugmentationsPreprint
Shen, H., Price, L. C., Bahadori, M. T., & Seeger, F. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.02.18.431877 · Cited by 16
Exploring the Protein Sequence Space with Global Generative ModelsPreprint
Romero‐Romero, S., Lindner, S., & Ferruz, N. (2023) · Cold Spring Harbor Perspectives in Biology · doi:10.1101/cshperspect.a041471 · Cited by 18
Li, Y.-F., Pan, X., & Shen, H.-B. (2025) · Patterns · doi:10.1016/j.patter.2025.101262 · Cited by 18
“14 Computational predictors combined with large language models have demonstrated superior performance compared with traditional methods in different fields of biology, such as structure prediction, 15 missense variant effect prediction, 16 peptide detection, 17 and subcellular localization…”
Cascade Catalytic Systems for Converting CO2 into C2+ Products
Shi, Q., Zhang, B., Wu, Z., Yang, D., Wu, H., Shi, J., & Jiang, Z. (2024) · ChemSusChem · doi:10.1002/cssc.202401916 · Cited by 14
Multi-dimensional metagenomics
Peng, H., Ruiz-Moreno, A. J., & Fu, J. (2025) · Nature Reviews Bioengineering · doi:10.1038/s44222-025-00346-x · Cited by 19
Evolutionary Context-Integrated Deep Sequence Modeling for Protein Engineering
Luo, Y., Vo, L. T., Ding, H., Su, Y., Liu, Y., Qian, W. W., Zhao, H., & Peng, J. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-45257-5_30 · Cited by 15
Construction of L-Asparaginase Stable Mutation for the Application in Food Acrylamide Mitigation
Yuan, B., Ma, P., Fan, Y., Guan, B., Hu, Y., Zhang, Y., Yan, W., Li, X., & Ni, Y. (2022) · Fermentation · doi:10.3390/fermentation8050218 · Cited by 14
High-Temperature Tolerance Protein Engineering through Deep Evolution
Chu, H., Tian, Z., Hu, L., Zhang, H., Chang, H., Bai, J., Liu, D., Lu, L., Cheng, J., & Jiang, H. (2024) · BioDesign Research · doi:10.34133/bdr.0031 · Cited by 14
“In addition, general protein language models, such as UniRep [31] and the Evolutionary Scale Modeling (ESM) [32], can embed the protein sequences into a high-dimensional representation space, in which it is more feasible to establish connections between protein properties and sequence variants…”
MethodExploring laccase: a sustainable enzymatic solution for the paper recycling domain
Pandey, S., & Gupta, S. (2024) · Archives of Microbiology · doi:10.1007/s00203-024-03927-3 · Cited by 14
“Its tremendous role is of great significance, aiming the translational approach (Alley et al. 2019; Song et al. 2023).”
Alcantar, M. A., English, M. A., Valeri, J. A., & Collins, J. J. (2024) · Molecular Cell · doi:10.1016/j.molcel.2024.05.025 · Cited by 15
Programmable biology through artificial intelligence: from nucleic acids to proteins to cells
Abudayyeh, O. O., & Gootenberg, J. S. (2024) · Nature Methods · doi:10.1038/s41592-024-02338-y · Cited by 15
Sequence, Structure, and Functional Space of Drosophila De Novo Proteins
Middendorf, L., Iyengar, B. R., & Eicholt, L. A. (2024) · Genome Biology and Evolution · doi:10.1093/gbe/evae176 · Cited by 15
Resources and Methods for Engineering “Designer” Glycan-Binding Proteins
Warkentin, R., & Kwan, D. H. (2021) · Molecules · doi:10.3390/molecules26020380 · Cited by 15
Breakthroughs in computational design methods open up new frontiers for de novo protein engineering
Meinen, B. A., & Bahl, C. D. (2021) · Protein Engineering Design and Selection · doi:10.1093/protein/gzab007 · Cited by 14
Performance of Regression Models as a Function of Experiment Noise
Li, G., Zrimec, J., Ji, B., Geng, J., Larsbrink, J., Zelezniak, A., Nielsen, J., & Engqvist, M. K. M. (2021) · Technical University of Denmark, DTU Orbit (Technical University of Denmark, DTU) · Cited by 14
TripletProt: Deep Representation Learning of Proteins Based On Siamese Networks
Nouranı, E., Asgari, E., McHardy, A. C., & Mofrad, M. R. K. (2021) · IEEE Transactions on Computational Biology and Bioinformatics · doi:10.1109/tcbb.2021.3108718 · Cited by 14
Herrington, N. B., & Kellogg, G. E. (2021) · Frontiers in Molecular Biosciences · doi:10.3389/fmolb.2021.773385 · Cited by 14
“, 2009) or even de novo structure prediction (Alley et al., 2019; Senior et al., 2020; Yang et al., 2020) involves many puzzle pieces and interactions, but some key information involving, e.”
BackgroundZagkou, S., Marais, V., Zeghoudi, N., Guillou, É. L., Eskelinen, E., Panasyuk, G., Verrier, B., & Primard, C. (2022) · Pharmaceutics · doi:10.3390/pharmaceutics14071379 · Cited by 15
“Modification of the peptide/protein structure [28], bioconjugation of peptides/proteins with polyethylene glycol (PEG) or PEG alternatives [29] and incorporation of the peptide/protein in drug delivery systems have been described [30–33].”
BackgroundProfile Prediction: An Alignment-Based Pre-Training Task for Protein Sequence ModelsPreprint
Sturmfels, P., Vig, J., Madani, A., & Rajani, N. F. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.00195 · Cited by 16
Filipavicius, M., Manica, M., Cadow, J., & Martínez, M. R. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.03084 · Cited by 25
“UniRep was the first study to apply deep learning to obtain protein representations from unlabeled sequences [2].”
BackgroundResultNow What Sequence? Pre-trained Ensembles for Bayesian Optimization of Protein SequencesPreprint
Yang, Z., Milas, K. A., & White, A. D. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.08.05.502972 · Cited by 15
“Alley et al. [3] used an mLSTM model to learn statistical representations of proteins from 24 million UniRef50 46 sequences.”
BackgroundMethodProteinGLUE multi-task benchmark suite for self-supervised protein modeling
Capel, H., Weiler, R., Dijkstra, M., Vleugels, R., Bloem, P., & Feenstra, K. A. (2022) · Scientific Reports · doi:10.1038/s41598-022-19608-4 · Cited by 11
“Several studies have shown correlation between the embeddings and physicochemical properties 11,12,36,37 , the source organism proteome 11,12,36,37 , the secondary structure 11,36,37 , and the subcellular location 11,37 .”
BackgroundMethodInfluentialLanguage models for protein design
Lee, J. S., Abdin, O., & Kim, P. M. (2025) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2025.103027 · Cited by 15
Xu, S., & Onoda, A. (2025) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.4c01288 · Cited by 13
Yu, H., Wang, Y., Yang, Z., Ying, J., Guan, F., Liu, B., Miao, M., Mohamed, A., Wei, X., Yang, Y., et al. (2025) · Food Chemistry · doi:10.1016/j.foodchem.2025.144208 · Cited by 13
Machine-learning-assisted universal protein activation in living mice
Wang, X., Liu, Y., Wang, Z., Zeng, X., Ngai, W. S. C., Wang, J., Zhang, H., Xie, X., Zhu, R., Fan, X., et al. (2025) · Cell · doi:10.1016/j.cell.2025.05.006 · Cited by 14
Relation Between the Number of Peaks and the Number of Reciprocal Sign Epistatic Interactions
Saona, R., Kondrashov, F. A., & Khudiakova, K. A. (2022) · Bulletin of Mathematical Biology · doi:10.1007/s11538-022-01029-z · Cited by 13
Predicting Protein Functions Based on Heterogeneous Graph Attention Technique
Zhao, Y., Yang, Z., Wang, L., Zhang, Y., Lin, H., & Wang, J. (2024) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2024.3357834 · Cited by 13
“Some approaches leveraging pre-trained language models have successfully captured the biological characteristics of sequences and significantly improved protein function prediction [9], [10], [11], [12].”
BackgroundFungtion: A Server for Predicting and Visualizing Fungal Effector Proteins
Li, J., Ren, J., Dai, W., Stubenrauch, C. J., Finn, R. D., & Wang, J. (2024) · Journal of Molecular Biology · doi:10.1016/j.jmb.2024.168613 · Cited by 13
Protein transfer learning improves identification of heat shock protein families
Min, S., Kim, H., Lee, B., & Yoon, S. (2021) · PLoS ONE · doi:10.1371/journal.pone.0251865 · Cited by 13
Unsupervised Representation Learning for Proteochemometric Modeling
Kim, P. T., Winter, R., & Clevert, D.-A. (2021) · International Journal of Molecular Sciences · doi:10.3390/ijms222312882 · Cited by 19
“Additionally, the finding that MolBert + UniRep is a good descriptor combination to use for the protein–ligand binding prediction task is replicated, as we find that this combination significantly outperforms all other combinations, except for MolBert + ESM (Table 3).”
MethodInfluentialHow Deep Learning Tools Can Help Protein Engineers Find Good Sequences
Osadchy, M., & Kolodny, R. (2021) · The Journal of Physical Chemistry B · doi:10.1021/acs.jpcb.1c02449 · Cited by 16
Deep representation learning improves prediction of LacI-mediated transcriptional repression
Garruss, A. S., Collins, K. M., & Church, G. M. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2022838118 · Cited by 14
“Such models, for example, have been shown to develop automatic pattern detection of protein secondary structure (29), thermostability (30), and localization (31).”
BackgroundMethodInfluentialSelf-citationMansoor, S., Baek, M., Madan, U., & Horvitz, E. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.09.01.458592 · Cited by 12
“Most methods that seek to understand the complex sequence to structure to function relationship take either a first principles approach with structural simulations [1] or leverage sequence embeddings trained through an adaptation of semi-supervised machine learning methods used to construct large-scale neural models d…”
MethodChen, J. Z., Gall, B., Pulsford, S. B., Tokuriki, N., & Jackson, C. J. (2025) · Molecular biology and evolution · doi:10.1093/molbev/msaf136 · Cited by 8
“…alternatives to the traditional homology-based neighborhoods in SSNs by using sequence vector representations from protein language models (pLM) (Alley et al. 2019; Madani et al. 2023; Rives et al. 2021) for sequence space exploration, and find them to perform comparably to the homology-based…”
Intelligent enzyme design and hybrid catalytic systems: Driving innovation in biocatalysis
Qu, G., Zhong, F., Xu, J. F., Huang, X., & Sun, Z. (2025) · Green Synthesis and Catalysis · doi:10.1016/j.gresc.2025.12.002 · Cited by 14
Zhuang, Y., Liu, X., Zhong, Y., & Wu, L. (2021) · IEEE Transactions on Computational Biology and Bioinformatics · doi:10.1109/tcbb.2021.3068381 · Cited by 11
Jiang, S., Zhang, S., Kang, X., Feng, Y., Li, Y., Nie, M., Li, Y., Chen, Y., Zhao, S., Jiang, T., et al. (2023) · Viruses · doi:10.3390/v15071556 · Cited by 12
“Deep learning based on unlabeled amino acids (AAs) was utilized to extract statistical representation with rich semantics from fundamental features of a protein [48].”
MethodKuang, Z., Zhao, Y., & Yang, X. (2023) · Agriculture Communications · doi:10.1016/j.agrcom.2023.100014 · Cited by 12
Jiang, J., Pei, H., Li, J., Li, M., Zou, Q., & Lv, Z. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae037 · Cited by 12
“Alley et al. [46] developed a deep learning-based sequence representation method, UniRep, a one-way mLSTM with 1900 hidden features.”
MethodA systematic analysis of regression models for protein engineering
Richard, M., Kæstel‐Hansen, J., Groth, P. M., Bartels, S., Salomon, J., Tian, P., Hatzakis, N. S., & Boomsma, W. (2024) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1012061 · Cited by 21
Improving viral annotation with artificial intelligence
Flamholz, Z., Li, C., & Kelly, L. (2024) · mBio · doi:10.1128/mbio.03206-23 · Cited by 12
“These self-supervised models, trained on millions of protein sequences, could resolve protein structural and functional information from only sequence input (56–61).”
BackgroundMethodAdvances in Zero‐Shot Prediction‐Guided Enzyme Engineering Using Machine Learning
Liu, C., Wu, J., Chen, Y., Liu, Y., Zheng, Y., Liu, L., & Zhao, J. B. (2024) · ChemCatChem · doi:10.1002/cctc.202401542 · Cited by 15
UniAMP: enhancing AMP prediction using deep neural networks with inferred information of peptides
Chen, Z., Ji, C., Xu, W., Gao, J., Huang, J., Xu, H., Qian, G., & Huang, J. (2025) · BMC Bioinformatics · doi:10.1186/s12859-025-06033-3 · Cited by 14
“Thirdly, the feature vectors of the peptides were computed using UniRep [25], ESM-2 [26], and ProtT5 [27], resulting in three vectors of 1900, 1280, and 1024 dimensions, respectively.”
Kilinc, M., Jia, K., & Jernigan, R. L. (2025) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2025.102984 · Cited by 15
Machine-Guided Dual-Objective Protein Engineering for Deimmunization and Therapeutic Functions
Wolfsberg, E., Paul, J.-S., Tycko, J., Chen, B., Bassik, M. C., Bintu, L., Alizadeh, A. A., & Gao, X. J. (2025) · Cell Systems · doi:10.1016/j.cels.2025.101299 · Cited by 14
Predicting the Disease Risk of Protein Mutation Sequences With Pre-training Model
Li, K., Zhong, Y., Lin, X., & Quan, Z. (2020) · Frontiers in Genetics · doi:10.3389/fgene.2020.605620 · Cited by 12
“For ProtVec, we generated the embedding of 300-dimensions for each sequence, and the sequences with terminators were treated in the same way as UniRep.”
BackgroundMethodInfluentialSPRoBERTa: protein embedding learning with local fragment modeling
Wu, L., Yin, C., Zhu, J., Wu, Z., He, L., Xia, Y., Xie, S., Qin, T., & Liu, T. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac401 · Cited by 12
“In this line,[6,7] are among the first works to apply deep learning to obtain protein representations from unlabeled protein sequences.”
BackgroundLim, H., & No, K. T. (2022) · BMC Bioinformatics · doi:10.1186/s12859-022-05010-4 · Cited by 12
“2 M parameters and 1900 features [33].”
BackgroundStructural biology at the scale of proteomes
Bouatta, N., & AlQuraishi, M. (2023) · Nature Structural & Molecular Biology · doi:10.1038/s41594-023-00924-w · Cited by 12
Self-citationIncorporating Pre-training Paradigm for Antibody Sequence-Structure Co-designPreprint
Gao, K., Wu, L., Zhu, J., Peng, T., Xia, Y., He, L., Xie, S., Qin, T., Liu, H., He, K., et al. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.11.14.516404 · Cited by 13
“Previously, people focus more on generating the CDR sequences only [30, 56, 3, 43].”
BackgroundBoosting the predictive power of protein representations with a corpus of text annotations
Duan, H., Skreta, M., Cotta, L., Rajaonson, E. M., Dhawan, N., Aspuru‐Guzik, A., & Maddison, C. J. (2025) · Nature Machine Intelligence · doi:10.1038/s42256-025-01088-6 · Cited by 11
Towards accurate artificial intelligence models for strain-level phage–host prediction
Malajczuk, C. J., Vaitekenas, A., Iszatt, J. J., Stick, S. M., Kicic, A., & Karpievitch, Y. V. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag085 · Cited by 11
“At the representation level, PBIP encodes protein sequences from both phages and bacterial hosts using the pretrained pLM UniRep [91].”
Molecular and Computational Strategies to Increase the Efficiency of CRISPR-Based Techniques
Mattiello, L., Rütgers, M., Sua-Rojas, M. F., Tavares, R., Soares, J. S., Begcy, K., & Menossi, M. (2022) · Frontiers in Plant Science · doi:10.3389/fpls.2022.868027 · Cited by 12
“Thereby, the use of conservative computational approaches such as UniRep (Alley et al., 2019), low-N (Biswas et al., 2021), and 1D convolutional neural network (CNN) is gaining much ground nowadays to the study of large-scale conformational changes of Cas9 and to accelerate engineered proteins…”
MethodUsing genetic programming to predict and optimize protein function
Miralavy, I., Bricco, A. R., Gilad, A. A., & Banzhaf, W. (2022) · PeerJ Physical Chemistry · doi:10.7717/peerj-pchem.24 · Cited by 11
Protein engineering via sequence-performance mapping
McConnell, A., & Hackel, B. J. (2023) · Cell Systems · doi:10.1016/j.cels.2023.06.009 · Cited by 13
A deep learning model for type II polyketide natural product prediction without sequence alignment
Huang, J., Gao, Q., Tang, Y., Wu, Y., Zhang, H., & Qin, Z. (2023) · Digital Discovery · doi:10.1039/d3dd00107e · Cited by 11
Qi, D., Chen, S., & Liu, T. (2024) · Analytical Biochemistry · doi:10.1016/j.ab.2024.115603 · Cited by 11
Zhang, L., Pang, H., Zhang, C., Li, S., Tan, Y., Jiang, F., Li, M., Yu, Y., Zhou, Z., Wu, B., et al. (2025) · Acta Pharmaceutica Sinica B · doi:10.1016/j.apsb.2025.03.028 · Cited by 14
Predictive and therapeutic applications of protein language models
Furui, K., Sakano, K., & Ohue, M. (2025) · Allergology International · doi:10.1016/j.alit.2025.08.004 · Cited by 12
Functional alignment of protein language models via reinforcement learning
Blalock, N., Seshadri, S., Nakamura, K., Babbar, A., Fahlberg, S. A., Kulkarni, A., & Romero, P. A. (2026) · Nature Communications · doi:10.1038/s41467-026-77557-2 · Cited by 19
Computational protein design with data‐driven approaches: Recent developments and perspectives
Liu, H., & Chen, Q. (2022) · Wiley Interdisciplinary Reviews Computational Molecular Science · doi:10.1002/wcms.1646 · Cited by 13
“Thus, representation-based models developed in the NLP field are being rapidly transferred to the protein domain, leading to several recently reported protein representation models including UniRep, 77 ESM, 78 TAPE (Tasks Assessing Protein Embeddings), 79 and ProtTrans.”
MethodPeng, W., & Sadaghiani, O. K. (2023) · Energy Sources Part A Recovery Utilization and Environmental Effects · doi:10.1080/15567036.2023.2232322 · Cited by 13
“(Alley et al. 2019) A comprehensive knowledge of protein function is necessary for rational protein engineering.”
BackgroundProtein language models meet reduced amino acid alphabets
Ieremie, I., Ewing, R. M., & Niranjan, M. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btae061 · Cited by 13
“We employ a three-layer BiLSTM similar to previous work (Bepler and Berger 2018, Alley et al. 2019) trained on masked token prediction.”
The applications of machine learning in HIV neutralizing antibodies research—A systematic review
Dănăilă, V.-R., Avram, S., & Buiu, C. (2022) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2022.102429 · Cited by 12
Removing bias in sequence models of protein fitnessPreprint
Shaw, A. Y., Spinner, H., Shin, J., Gurev, S. F., Rollins, N. J., & Marks, D. S. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.09.28.560044 · Cited by 13
Automated Scientific Discovery: From Equation Discovery to Autonomous Discovery Systems
Kramer, S., Cerrato, M., Brugger, J., Džeroski, S., & King, R. (2023) · arXiv · doi:10.48550/arxiv.2305.02251 · Cited by 10
“In all the existing AI Scientists systems that we are aware off the type of experiment that can be executed are limited to a small stereotypical set.”
PRESCOTT: a population aware, epistatic, and structural model accurately predicts missense effects
Tekpinar, M., David, L., Henry, T., & Carbone, A. (2025) · Genome Biology · doi:10.1186/s13059-025-03581-y · Cited by 18
Fine-tuning protein language models unlocks the potential of underrepresented viral proteomes
Sawhney, R., Ferrell, B. D., Dejean, T., Schreiber, Z., Harrigan, W., Polson, S. W., Wommack, K. E., & Belcaid, M. (2025) · PeerJ · doi:10.7717/peerj.19919 · Cited by 14
Machine learning for enzyme catalytic activity: current progress and future horizons
Qiu, S., Saeed, H., Leonard, W., Li, F., & Yang, A. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag002 · Cited by 13
“In the comparison of different sequence encoding methods, transfer learning from OGTs outcompeted iFeatures [106] and UniRep [107].”
Wu, Y. F., Chen, K.-W., Li, Y., Qiu, K., Lu, G., Shu, T., Sun, T., Hou, Z., Wu, Y., & Yu, L. (2023) · Journal of Agricultural and Food Chemistry · doi:10.1021/acs.jafc.3c03431 · Cited by 10
Case, M., Smith, M., Vinh, J., & Thurber, G. M. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2311726121 · Cited by 13
Molecular sequence classification using efficient kernel based embedding
Ali, S., Ali, T. E., Murad, T., Mansoor, H., & Patterson, M. (2024) · Information Sciences · doi:10.1016/j.ins.2024.121100 · Cited by 11
Zhang, G., Wang, H., Zhang, Z., Zhang, L., Guo, G., Yang, J., Yuan, F., & Ju, F. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae319 · Cited by 12
“…(i.e. FunTrans and FunRep) in FunGeneTyper, which can learn more hidden features of protein sequences, especially the context information [19, 21], compared to the traditional one-hot encoding algorithm and the convolutional neural network used by HMD-ARG [35] and the multilayer perceptron…”
BackgroundYönden, Z., Reshadi, S., Hayati, A. F., Hooshiar, M. H., Ghasemi, S., Yönden, H., & Daemi, A. (2025) · Drug Development Research · doi:10.1002/ddr.70066 · Cited by 10
“By training RNNs on amino acid sequences, researchers can create embedded representations of proteins that can be used to predict various properties, such as protein structure and function (Alley et al. 2019).”
In silico prediction of variant effects: promises and limitations for precision plant breeding
Sendrowski, J., Bataillon, T. M., & Ramstein, G. P. (2025) · Theoretical and Applied Genetics · doi:10.1007/s00122-025-04973-1 · Cited by 12
“…aim to predict conservation by taking into account the sequence context of the focal locus, either without incorporating alignment information (Alley et al. 2019; Rives et al. 2021; Elnaggar et al. 2022; Benegas et al. 2023; Mendoza-Revilla et al. 2024; Zhai et al. 2025) or with it (Rao et…”
ProSE: the architecture and design of a protein discovery engine
Robson, E. S., Xu, C., & Wills, L. W. (2022) · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) · doi:10.1145/3503222.3507722 · Cited by 10
“Substantial groundbreaking work has documented the success of NLP models in protein modeling [1, 3, 8, 35, 43, 45], especially those employing BERT-style structures [11].”
PROVAL: A framework for comparison of protein sequence embeddings
Väth, P., Münch, M., Raab, C., & Schleif, F.-M. (2022) · Journal of Computational Mathematics and Data Science · doi:10.1016/j.jcmds.2022.100044 · Cited by 11
Wang, Q., Wei, J. S., Zhou, Y., Lin, M., Ren, R., Wang, S., Cui, S. R., & Li, Z. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac351 · Cited by 12
Identification of plant vacuole proteins by exploiting deep representation learning features
Jiao, S., & Zou, Q. (2022) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2022.06.002 · Cited by 10
“Typical examples of deep learning protein sequence embedding methods include UniRep [28], bidirectional long short-term memory embedding model (BiLSTM) [29], TAPE [30], SSA [14] and language embedding model (LM) [31].”
InfluentialImproving protein structure prediction using templates and sequence embedding
Wu, F., Jing, X., Luo, X., & Xu, J. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac723 · Cited by 12
learnMSA: learning and aligning large protein families
Becker, F., & Stanke, M. (2022) · GigaScience · doi:10.1093/gigascience/giac104 · Cited by 9
“The field of protein language modeling where parameter-rich sequence models are learned semisupervised [44, 45] based on Attention [46, 47] or LSTMs [48] is also compatible and complementary to our approach.”
BackgroundResultJohn, C., Sahoo, J., Madhavan, M., & Mathew, O. K. (2023) · Current Bioinformatics · doi:10.2174/1574893618666230320103421 · Cited by 14
Ong, W. S. Y., Ji, K., Pathiranage, V., Maydew, C., Baek, K., Villones, R. L. E., Meloni, G., Walker, A. R., & Dodani, S. C. (2023) · Angewandte Chemie International Edition · doi:10.1002/anie.202302304 · Cited by 10
Intelligent Protein Design and Molecular Characterization Techniques: A Comprehensive Review
Wang, J., Chen, C., Yao, G., Ding, J., Wang, L., & Jiang, H. (2023) · Molecules · doi:10.3390/molecules28237865 · Cited by 10
“NLP-based learning models for protein sequence representation have achieved remarkable results in protein design [34,37,161,178,179].”
Background
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
