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 301–400 of 1,003
Jin, S., Zeng, Z., Xiong, X., Huang, B., Tang, L., Wang, H., Ma, X., Tang, X., Shao, G., Huang, X., et al. (2025) · Communications Biology · doi:10.1038/s42003-025-08282-7 · Cited by 43
Li, G., Qin, Y., Fontaine, N., Chong, M. N. F., Maria‐Solano, M. A., Feixas, F., Cadet, X. F., Pandjaitan, R., Garcia‐Borràs, M., Cadet, F., et al. (2020) · ChemBioChem · doi:10.1002/cbic.202000612 · Cited by 37
Kandathil, S. M., Greener, J. G., Lau, A. M., & Jones, D. T. (2022) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2113348119 · Cited by 48
Protein Design and Variant Prediction Using Autoregressive Generative ModelsPreprint
Shin, J.-E., Riesselman, A. J., Kollasch, A. W., McMahon, C., Simon, E. P., Sander, C., Manglik, A., Kruse, A. C., & Marks, D. S. (2019) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/757252 · Cited by 40
Li, G., Yao, S., & Fan, L. (2024) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.3c01697 · Cited by 37
AI for organic and polymer synthesis
Hong, X., Yang, Q., Liao, K., Pei, J., Chen, M., Mo, F., Lu, H., Zhang, W.-B., Zhou, H., Chen, J., et al. (2024) · Science China Chemistry · doi:10.1007/s11426-024-2072-4 · Cited by 41
Automatic Gene Function Prediction in the 2020’s
Makrodimitris, S., van Ham, R. C. H. J., & Reinders, M. J. T. (2020) · Genes · doi:10.3390/genes11111264 · Cited by 38
A dataset comprised of binding interactions for 104,972 antibodies against a SARS-CoV-2 peptide
Engelhart, E., Emerson, R., Shing, L., Lennartz, C., Guion, D., Kelley, M., Lin, C., Lopez, R. M., Younger, D. A., & Walsh, M. (2022) · Scientific Data · doi:10.1038/s41597-022-01779-4 · Cited by 37
“Phage display, for example, is intended to provide the researcher with information on a small number of the top binders from a pool of > 10 6 designs 14,15 .”
BackgroundMachine Learning: A Suitable Method for Biocatalysis
Sampaio, P., & Fernandes, P. C. D. B. (2023) · Catalysts · doi:10.3390/catal13060961 · Cited by 39
“ANNs: modelling and optimization of enzymatic treatment for nutritional enhancement of rice [140], optimization of fermentation conditions for production of lipopeptide antibiotic [141], optimization of algal biofuel production [20], prediction of the toxicity of ionic liquids towards enzyme activity [142], liquid lev…”
MethodDomain loss enabled evolution of novel functions in the snake three-finger toxin gene superfamily
Koludarov, I., Senoner, T., Jackson, T., Dashevsky, D., Heinzinger, M., Aird, S. D., & Rost, B. (2023) · Nature Communications · doi:10.1038/s41467-023-40550-0 · Cited by 38
Gong, J., Jiang, L., Chen, Y., Zhang, Y., Li, X., Ma, Z., Fu, Z., He, F., Sun, P., Ren, Z., et al. (2023) · Bioinformatics · doi:10.1093/bioinformatics/btad646 · Cited by 37
“…protein language models (PLMs), including ESM-1b (Rives et al. 2021), ESM-2 (Lin et al. 2023), ProtBert (Elnaggar et al. 2022), and UniRep (Alley et al. 2019), attention-based architectures, which can effectively extract contextual information from sequences, have been adopted to extract…”
MethodRapid directed evolution guided by protein language models and epistatic interactions
Tran, V. Q., Nemeth, M., Bartie, L. J., Chandrasekaran, S. S., Fanton, A., Moon, H. C., Hie, B., Konermann, S., & Hsu, P. D. (2026) · Science · doi:10.1126/science.aea1820 · Cited by 41
PyPEF—An Integrated Framework for Data-Driven Protein Engineering
Siedhoff, N. E., Illig, A.-M., Schwaneberg, U., & Davari, M. D. (2021) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.1c00099 · Cited by 36
Li, M., Kang, L., Xiong, Y., Wang, Y. G., Fan, G., Tan, P., & Hong, L. (2023) · Journal of Cheminformatics · doi:10.1186/s13321-023-00688-x · Cited by 38
“…from evolution-related sequences[15, 16]; the second one (global encoder) captures the sequence feature in global protein sequence universe[6, 12]; and the third one (structure module) captures local structural feature around each residue learned from 3D geometric structure of the…”
BackgroundMethodMalbranke, C., Bikard, D., Cocco, S., Monasson, R., & Tubiana, J. (2023) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2023.102571 · Cited by 40
Ramstein, G. P., & Buckler, E. S. (2022) · Genome Biology · doi:10.1186/s13059-022-02747-2 · Cited by 34
Dong, T. N., Brogden, G., Gerold, G., & Khosla, M. (2021) · BMC Bioinformatics · doi:10.1186/s12859-021-04484-y · Cited by 39
Rai, B. K., Apgar, J. R., & Bennett, E. M. (2023) · Scientific Reports · doi:10.1038/s41598-023-28841-4 · Cited by 34
Palit, P., Minkara, M., Abida, M., Marwa, S., Sen, C., Roy, A., Pasha, M. R., Paulraj, M. S., Saha, A., & Ferdoush, J. (2025) · Biomolecules · doi:10.3390/biom15050684 · Cited by 35
“AI-powered computational platforms can predict the effects of specific gene edits on enzyme function and stability using molecular dynamics simulations and structure–function modeling [35].”
Pretraining model for biological sequence data
Song, B., Li, Z., Lin, X., Wang, J., Wang, T., & Fu, X. (2021) · Briefings in Functional Genomics · doi:10.1093/bfgp/elab025 · Cited by 37
“[66] build unified representation (UniRep) from massive unlabeled amino acid sequences by a multi-LSTM model.”
BackgroundGautam, V., Gaurav, A., Masand, N., Lee, V. S., & Patil, V. M. (2022) · Molecular Diversity · doi:10.1007/s11030-022-10489-3 · Cited by 40
Pei, H., Li, J., Ma, S., Jiang, J., Li, M., Zou, Q., & Lv, Z. (2023) · Applied Sciences · doi:10.3390/app13052858 · Cited by 38
Towards new frontiers of healthcare systems research using artificial intelligence and generative AI
Chatterjee, S., Fruhling, A., Kotiadis, K., & Gärtner, D. (2024) · Health Systems · doi:10.1080/20476965.2024.2402128 · Cited by 42
“AlphaFold represented a breakthrough in the key task of protein folding (Senior et al., 2020), while Alley et al. made strides in the area of protein analysis, creating statistical summaries that help neural networks learn with less data (Alley et al., 2019).”
Sun, J., Zhu, T., Cui, Y., & Wu, B. (2025) · The Innovation · doi:10.1016/j.xinn.2024.100750 · Cited by 45
Conditional generative modeling for de novo protein design with hierarchical functions
Kucera, T., Togninalli, M., & Meng-Papaxanthos, L. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac353 · Cited by 38
Pre-trained Language Models in Biomedical Domain: A Systematic SurveyPreprint
Wang, B., Xie, Q., Pei, J., Chen, Z., Tiwari, P., Li, Z., & Fu, J. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2110.05006 · Cited by 34
Hu, F., Hu, Y., Zhang, W., Huang, H., Pan, Y. F., & Yin, P. (2023) · Advanced Science · doi:10.1002/advs.202301223 · Cited by 33
Enhanced Sequence-Activity Mapping and Evolution of Artificial Metalloenzymes by Active Learning
Vornholt, T., Mutný, M., Schmidt, G. W., Schellhaas, C., Tachibana, R., Panke, S., Ward, T. R., Krause, A., & Jeschek, M. (2024) · ACS Central Science · doi:10.1021/acscentsci.4c00258 · Cited by 36
“Furthermore, we developed a machine learning model based on Gaussian process regression that incorporates optimized descriptors and estimates of experimental noise to efficiently navigate the sequence space.”
MethodInfluentialA machine learning toolkit for genetic engineering attribution to facilitate biosecurity
Alley, E. C., Turpin, M., Liu, A. B., Kulp-McDowall, T., Swett, J. L., Edison, R., Stetina, S. E. V., Church, G. M., & Esvelt, K. M. (2020) · Nature Communications · doi:10.1038/s41467-020-19612-0 · Cited by 35
Self-citationBioengineered Protein‐based Adhesives for Biomedical Applications
Sun, J., Han, J., Wang, F., Liu, K., & Zhang, H. (2021) · Chemistry - A European Journal · doi:10.1002/chem.202102902 · Cited by 33
Large-scale design and refinement of stable proteins using sequence-only models
Singer, J. M., Novotney, S., Strickland, D., Haddox, H. K., Leiby, N., Rocklin, G. J., Chow, C. M., Roy, A., Bera, A. K., Motta, F. C., et al. (2022) · PLoS ONE · doi:10.1371/journal.pone.0265020 · Cited by 34
“Alley et al. (2019) developed a sequence-based recurrent model which generated fixed-length vector embeddings of proteins, based solely on primary sequence.”
BackgroundGenerative AAV capsid diversification by latent interpolationPreprint
Sinai, S., Jain, N. K., Church, G. M., & Kelsic, E. D. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.04.16.440236 · Cited by 39
“However, more recently, data-driven machine learning efforts for protein engineering have seen a rapid uptick, targeting protein structure and (to a lesser degree) function prediction (Alley et al., 2019; AlQuraishi, 2019; Hiranuma et al., 2021; Hopf et al., 2017; Marks et al., 2011; Norn et al., 2021; Rao et al., 201…”
BackgroundSelf-citationSVSBI: sequence-based virtual screening of biomolecular interactions
Shen, L., Feng, H., Qiu, Y., & Wei, G. (2023) · Communications Biology · doi:10.1038/s42003-023-04866-3 · Cited by 32
Zhang, Y., Guan, J., Li, C., Wang, Z., Deng, Z., Gasser, R. B., Song, J., & Ou, H. (2023) · Research · doi:10.34133/research.0258 · Cited by 39
“These representations are reported to reflect structure, evolutionary, and/or biophysical contexts [34].”
BackgroundChaisupa, P., & Wright, R. C. (2023) · SLAS TECHNOLOGY · doi:10.1016/j.slast.2023.10.005 · Cited by 39
Bashour, H., Smorodina, E., Pariset, M., Zhong, J., Akbar, R., Chernigovskaya, M., Lê Quý, K., Snapkow, I., Rawat, P., Krawczyk, K., et al. (2024) · Communications Biology · doi:10.1038/s42003-024-06561-3 · Cited by 32
“(ii ) The second embedding is an amino acid sequence encoding output produced by the protein language model (PLM) ESM-1v, which generates large semantically-rich digital representations of antibodies (high-dimensional embedding — order of magnitude ≈ 10 3 ) 74 – 76 .”
Is Transfer Learning Necessary for Protein Landscape Prediction?Preprint
Shanehsazzadeh, A., Belanger, D., & Dohan, D. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.03443 · Cited by 62
“We are unaware of other directly comparable works that use the same datasets, but machine learning techniques have been used in a similar fashion to learn useful protein embeddings in [1, 2, 13, 17, 17, 18].”
BackgroundResultSimultaneous enhancement of multiple functional properties using evolution-informed protein design
Fram, B., Su, Y., Truebridge, I., Riesselman, A. J., Ingraham, J. B., Passera, A., Napier, E., Thadani, N. N., Lim, S., Roberts, K., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-49119-x · Cited by 34
Domain-PFP allows protein function prediction using function-aware domain embedding representations
Ibtehaz, N., Kagaya, Y., & Kihara, D. (2023) · Communications Biology · doi:10.1038/s42003-023-05476-9 · Cited by 31
Shen, J., Yu, Q., Chen, S., Tan, Q., Li, J., & Li, Y. (2023) · Nature Computational Science · doi:10.1038/s43588-023-00576-2 · Cited by 32
“These protein language models [25, 26] gain improvement on problems with limited annotated data.”
Advancing bioinformatics with language models: components, applications, and perspectives
Liu, J., Yang, M., Yu, Y., Xu, H., Wang, T., Li, K., & Zhou, X. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag367 · Cited by 66
EITLEM-Kinetics: A deep-learning framework for kinetic parameter prediction of mutant enzymes
Shen, X., Cui, Z., Long, J., Zhang, S., Chen, B., & Tan, T. (2024) · Chem Catalysis · doi:10.1016/j.checat.2024.101094 · Cited by 37
Boob, A. G., Tan, S.-I., Zaidi, A., Singh, N., Xue, X., Zhou, S., Martin, T. A., Chen, L.-Q., & Zhao, H. (2025) · Nature Communications · doi:10.1038/s41467-025-59499-3 · Cited by 31
“We calculated the Levenshtein distance to determine the number of mutations between the generated sequence and its closest match.”
Anteghini, M., dos Santos, V. A. P. M., & Saccenti, E. (2021) · International Journal of Molecular Sciences · doi:10.3390/ijms22126409 · Cited by 30
“Two of the most promising DL-embeddings are the Unified Representation (UniRep) embedding [28] and the Sequence-to-Vector (SeqVec) embedding [31].”
BackgroundMethodInfluentialPeter, S. B., Qiao, Z., Godspower, H. N., Ajeje, S. B., Xu, M., Zhang, X., Yang, T., & Rao, Z. (2022) · Frontiers in Bioengineering and Biotechnology · doi:10.3389/fbioe.2021.802031 · Cited by 32
Wang, Y., Luo, X., & Zou, Q. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac374 · Cited by 31
High-throughput deep learning variant effect prediction with Sequence UNET
Dunham, A. S., Beltrao, P., & AlQuraishi, M. (2023) · Genome Biology · doi:10.1186/s13059-023-02948-3 · Cited by 31
Self-citationBeyond sequence: Structure-based machine learning
Durairaj, J., de Ridder, D., & van Dijk, A. D. J. (2022) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2022.12.039 · Cited by 32
DeepMineLys: Deep mining of phage lysins from human microbiome
Fu, Y., Yu, S., Li, J., Lao, Z., Yang, X., & Lin, Z. (2024) · Cell Reports · doi:10.1016/j.celrep.2024.114583 · Cited by 31
Structure-informed protein language models are robust predictors for variant effects
(2024) · doi:10.1007/s00439-024-02695-w · Cited by 30
“…across protein families that leverage alignments during training, such as the MSA Transformer (Rao et al. 2021) or that are alignment-free, such as UniRep (Alley et al. 2019), TAPE (Rao et al. 2019), ESM-1b (Rives et al. 2019), ESM-1v (Meier et al. 2021) and ProtBERT-BFD (Elnaggar et al. 2021).”
InfluentialIdentify Bitter Peptides by Using Deep Representation Learning Features
Jiang, J., Lin, X., Jiang, Y., Jiang, L., & Lv, Z. (2022) · International Journal of Molecular Sciences · doi:10.3390/ijms23147877 · Cited by 30
“Feature Extraction To explore the effects of different features on bitter peptide recognition, we used three deep representation learning feature extraction methods, namely, SSA [30], UniRep [25], and BiLSTM [31].”
MethodMedium-sized protein language models perform well at transfer learning on realistic datasets
Vieira, L. C., Handojo, M. L., & Wilke, C. O. (2025) · Scientific Reports · doi:10.1038/s41598-025-05674-x · Cited by 41
HNetGO: protein function prediction via heterogeneous network transformer
Zhang, X., Guo, H., Zhang, F., Wang, X., Wu, K., Qiu, S., Liu, B., Wang, Y., Hu, Y., & Li, J. (2021) · Briefings in Bioinformatics · doi:10.1093/bib/bbab556 · Cited by 30
Matsunaga, R., & Tsumoto, K. (2025) · Journal of Biomedical Science · doi:10.1186/s12929-025-01141-x · Cited by 30
“It can generate a 1900-dimensional embedding for any given protein sequence [53], providing valuable information for protein engineering tasks, such as predicting binding affinity, stability, and expression levels.”
Mughram, M. H. A., Catalano, C., Bowry, J. P., Safo, M. K., Scarsdale, J. N., & Kellogg, G. E. (2021) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.1c00235 · Cited by 28
Finding functional motifs in protein sequences with deep learning and natural language models
Savojardo, C., Martelli, P. L., & Casadio, R. (2023) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2023.102641 · Cited by 29
Mutexa: A Computational Ecosystem for Intelligent Protein Engineering
Yang, Z. J., Shao, Q., Jiang, Y., Jurich, C., Ran, X., Juarez, R. J., Yan, B., Stull, S. L., Gollu, A., & Ding, N. (2023) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.3c00602 · Cited by 28
Wei, L., Li, Q., Song, Y., Stefanov, S., Dong, R., Fu, N., Siriwardane, E. M. D., Chen, F., & Hu, J. (2024) · Advanced Science · doi:10.1002/advs.202304305 · Cited by 43
Feehan, R., Montezano, D., & Slusky, J. S. G. (2021) · Protein Engineering Design and Selection · doi:10.1093/protein/gzab019 · Cited by 29
Upadhyay, V., Boorla, V. S., & Maranas, C. D. (2023) · Metabolic Engineering · doi:10.1016/j.ymben.2023.06.001 · Cited by 26
Energy-based models for atomic-resolution protein conformationsPreprint
Du, Y., Meier, J., Ma, J., Fergus, R., & Rives, A. W. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.13167 · Cited by 35
Ultra-fast protein structure prediction to capture effects of sequence variation in mutation moviesPreprint
Weißenow, K., Heinzinger, M., Steinegger, M., & Rost, B. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.11.14.516473 · Cited by 29
Design in the DARK: Learning Deep Generative Models for De Novo Protein Design
Moffat, L., Kandathil, S. M., & Jones, D. T. (2022) · doi:10.1101/2022.01.27.478087 · Cited by 25
“Other examples are recent work using language models to learn representations for downstream supervised tasks (Alley et al., 2019; Bepler & Berger, 2019; Elnaggar et al., 2020; Rao et al., 2021).”
BackgroundMethodRecent Advances in Peptide Drug Discovery: Novel Strategies and Targeted Protein Degradation
Vrbnjak, K., & Sewduth, R. (2024) · Pharmaceutics · doi:10.3390/pharmaceutics16111486 · Cited by 24
“Large language models, in particular, have been beneficial in this field, and protein language models have been developed to study the function and structure of proteins based only on their amino acid sequences [36,37].”
Li, J., Wu, Y., Zhang, S., Kang, X., & Jiang, T. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac036 · Cited by 28
“Natural language processing methods have recently been shown to perform well in genomic sequence embedding [42, 43], particularly in the identification of contextual meaning in viral genomes.”
Shi, H., & Zhang, S. (2022) · Interdisciplinary Sciences Computational Life Sciences · doi:10.1007/s12539-022-00521-3 · Cited by 27
“Recently, Zhuang et al. [19] proposed an idea to capture the rich semantic attributes of AHTPs by using the existing pretraining coding algorithm [20] and convolutional neural network.”
MethodBoeynaems, S., Ma, X. R., Yeong, V., Ginell, G. M., Chen, J.-H., Blum, J. A., Nakayama, L., Sanyal, A., Briner, A., Haver, D. V., et al. (2023) · SSRN Electronic Journal · doi:10.1101/2023.03.09.531820 · Cited by 24
Deep learning methods for 3D structural proteome and interactome modeling
Lee, D., Xiong, D., Wierbowski, S., Li, L., Liang, S., & Yu, H. (2022) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2022.102329 · Cited by 26
“UniRep [58] is a Multiplicative LSTM (mLSTM)-based model that takes in unlabeled protein sequences and distills the fundamental features of the input into statistical representations.”
BackgroundFrom traditional to AI-driven: The evolution of intelligent enzyme engineering for biocatalysis
Lv, F., Zhang, J., You, S., & Qi, W. (2025) · Biotechnology Advances · doi:10.1016/j.biotechadv.2025.108788 · Cited by 31
Clustering FunFams using sequence embeddings improves EC purity
Littmann, M., Bordin, N., Heinzinger, M., Schütze, K., Dallago, C., Orengo, C. A., & Rost, B. (2021) · Bioinformatics · doi:10.1093/bioinformatics/btab371 · Cited by 27
“This simple global average pooling provides 11 an effective baseline [12, 15, 20].”
InfluentialDiscrete and continuous representations and processing in deep learning: Looking forward
Cartuyvels, R., Spinks, G., & Moens, M. (2021) · AI Open · doi:10.1016/j.aiopen.2021.07.002 · Cited by 32
“…et al., 2015; Chandak et al., 2019), distributed representations for atoms (Schütt et al., 2018), sequences of amino-acids in proteins (Alley et al., 2019), social network posts (Veli ˇ ckovi ´ c et al., 2019), tactile experience or touch (Gao et al., 2016; Yuan et al., 2017; Calandra…”
BackgroundBioinformatics approaches for unveiling virus-host interactions
Iuchi, H., Kawasaki, J., Kubo, K., Fukunaga, T., Hokao, K., Yokoyama, G., Ichinose, A., Suga, K., & Hamada, M. (2023) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2023.02.044 · Cited by 25
“extracts the latent representation of 24 million protein sequences [50] and fine-tuned this model via multi-task learning for both virus–human and human–human PPIs.”
Using protein language models for protein interaction hot spot prediction with limited data
Sargsyan, K., & Lim, C. (2024) · BMC Bioinformatics · doi:10.1186/s12859-024-05737-2 · Cited by 25
“Analogous to how these models decode human language, deep-learning models known as protein language models have emerged to decipher the intricate language of life by learning the patterns embedded in protein sequences over evolutionary time [2–6].”
BackgroundHeinzinger, M., & Rost, B. (2025) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2025.102986 · Cited by 27
Vahabi, L., Ranjbar, P. R., & Davar, F. (2022) · Journal of Drug Delivery Science and Technology · doi:10.1016/j.jddst.2022.104144 · Cited by 26
Ito, T., Nguyen, T. D., Saitō, Y., Kurumida, Y., Nakazawa, H., Kawada, S., Nishi, H., Tsuda, K., Kameda, T., & Umetsu, M. (2023) · mAbs · doi:10.1080/19420862.2023.2168470 · Cited by 27
Intrinsic-Extrinsic Convolution and Pooling for Learning on 3D Protein StructuresPreprint
Hermosilla, P., Schäfer, M., Lang, M., Fackelmann, G., Vázquez, P., Kozlíková, B., Krone, M., Ritschel, T., & Ropinski, T. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2007.06252 · Cited by 54
IgGM: A Generative Model for Functional Antibody and Nanobody DesignPreprint
Wang, R., Wu, F., Gao, X., Wu, J., Zhao, P., & Yao, J. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.09.19.613838 · Cited by 27
“Sequence models (Alley et al., 2019; Saka et al., 2021; Shin et al., 2021; Akbar et al., 2022) were widely used in the early stages of deep learning for antibody design.”
Protein Language Model Fitness Is a Matter of PreferencePreprint
Gordon, C., Lu, A. X., & Abbeel, P. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.10.03.616542 · Cited by 28
“We perform unsupervised finetuning on homologous sequences (sometimes referred to as evo-tuning (Alley et al., 2019; Biswas et al., 2021)) of ESM-2 650M on a set of sequences derived by searching UniRef100 for the wild type sequence on each of the 217 DMS studies in ProteinGym separately.”
MethodResultInfluentialData-driven enzyme engineering to identify function-enhancing enzymes
Jiang, Y., Ran, X., & Yang, Z. J. (2022) · Protein Engineering Design and Selection · doi:10.1093/protein/gzac009 · Cited by 27
Gu, J., Mu, W., Xu, Y., & Nie, Y. (2024) · Biotechnology Advances · doi:10.1016/j.biotechadv.2024.108496 · Cited by 27
AI-driven innovation in antibody-drug conjugate design
Noriega, H. A., & Wang, X. S. (2025) · Frontiers in Drug Discovery · doi:10.3389/fddsv.2025.1628789 · Cited by 26
CLADE 2.0: Evolution-Driven Cluster Learning-Assisted Directed Evolution
Qiu, Y., & Wei, G. (2022) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.2c01046 · Cited by 23
Machine learning methods for predicting protein structure from single sequences
Kandathil, S. M., Lau, A. M., & Jones, D. T. (2023) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2023.102627 · Cited by 26
“…had the capacity to predict multiple levels of features, such as amino acid-level physicochemistry, secondary structure, protein disorder, subcellular location, whether a protein was membrane-bound, as well as recognising the relationship between sequences in different model organisms [15,16].”
BackgroundMethodProtein language models can capture protein quaternary state
Avraham, O., Tsaban, T., Ben-Aharon, Z., Tsaban, L., & Schueler‐Furman, O. (2023) · BMC Bioinformatics · doi:10.1186/s12859-023-05549-w · Cited by 24
“1] solver = ['sgd', 'adam'] max_iter = [10, 100, 400, 800, 1000, 1500] n_iter_no_change = [5, 10, 15, 20, 30, 40] tol = [0.”
Best practices for machine learning in antibody discovery and development
Wossnig, L., Furtmann, N., Buchanan, A. G., Kumar, S., & Greiff, V. (2024) · Drug Discovery Today · doi:10.1016/j.drudis.2024.104025 · Cited by 25
ProteinNPT: Improving Protein Property Prediction and Design with Non-Parametric Transformers
Notin, P., Weitzman, R., Marks, D. S., & Gal, Y. (2023) · neural information processing systems · doi:10.52202/075280-1458 · Cited by 32
Combining evolutionary and assay-labelled data for protein fitness predictionPreprint
Hsu, C., Nisonoff, H. M., Fannjiang, C., & Listgarten, J. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.03.28.437402 · Cited by 24
Genome-wide prediction of disease variants with a deep protein language model
Brandes, N., Goldman, G., Wang, C. H., Ye, C. J., & Ntranos, V. (2022) · doi:10.1101/2022.08.25.505311 · Cited by 24
Design of microbial catalysts for two-stage processes
Shabestary, K., Klamt, S., Link, H., Mahadevan, R., Steuer, R., & Hudson, E. P. (2024) · Nature Reviews Bioengineering · doi:10.1038/s44222-024-00225-x · Cited by 23
“Computational tools can reduce the screening necessary to achieve a desired enhanced phenotype 215,216 (Fig.”
SPOT: A machine learning model that predicts specific substrates for transport proteins
Kroll, A., Niebuhr, N., Butler, G., & Lercher, M. J. (2024) · PLoS Biology · doi:10.1371/journal.pbio.3002807 · Cited by 29
A unified evolution-driven deep learning framework for virus variation driver prediction
Nie, Z., Liu, X., Chen, J., Wang, Z., Liu, Y., Si, H.-R., Dong, T., Xu, F., Song, G., Wang, Y., et al. (2025) · Nature Machine Intelligence · doi:10.1038/s42256-024-00966-9 · Cited by 25
Comprehensive benchmarking of large language models for RNA secondary structure prediction
Zablocki, L. I., Bugnon, L. A., Gerard, M. F., Persia, L. E. D., Stegmayer, G. S., & Milone, D. H. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf137 · Cited by 31
“Motivated by the success of large language models (LLM) in proteins [32, 37, 18, 3, 2, 39, 14, 31] and DNA [26], several RNA LLM have recently appeared [1, 36, 8, 55, 61, 58, 54] with the potential to be used for improving several important RNA-related tasks, among which one of the most relevant is…”
SidechainNet : An all‐atom protein structure dataset for machine learning
King, J. E., & Koes, D. R. (2021) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.26169 · Cited by 23
“In addition to the dominance of the AlphaFold deep learning method in the 2018 Critical Assessment of protein Structure Prediction (CASP) competition [1, 2, 3], many novel deep learning-based methods have been developed for protein representation [4, 5], property prediction [6, 7], and structure prediction [8, 9, 10,…”
MethodGalmés, M. À., Nödling, A. R., Luk, L. Y. P., Świderek, K., & Moliner, V. (2021) · ACS Catalysis · doi:10.1021/acscatal.1c02150 · Cited by 23
Burnim, A. A., Spence, M. A., Xu, D. Z., Jackson, C. J., & Ando, N. (2022) · eLife · doi:10.7554/elife.79790 · Cited by 25
Latent spaces for antimicrobial peptide design
Renaud, S., & Mansbach, R. A. (2023) · Digital Discovery · doi:10.1039/d2dd00091a · Cited by 27
Deep learning in preclinical antibody drug discovery and development
Zhou, Y., Huang, Z., Li, W., Wei, J., Jiang, Q., Yang, W., & Huang, J. H. (2023) · Methods · doi:10.1016/j.ymeth.2023.07.003 · Cited by 25
Rugged fitness landscapes minimize promiscuity in the evolution of transcriptional repressors
Meger, A., Spence, M. A., Sandhu, M., Matthews, D. S., Chen, J., Jackson, C. J., & Raman, S. (2024) · Cell Systems · doi:10.1016/j.cels.2024.03.002 · Cited by 23
FLAb: Benchmarking deep learning methods for antibody fitness predictionPreprint
Chungyoun, M., Ruffolo, J. A., & Gray, J. J. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.01.13.575504 · Cited by 27
“As the diversity of deep learning approaches increases [10, 29, 32, 31, 9, 18, 2, 3], it becomes vital to converge on a systematic benchmark for evaluating performance.”
BackgroundMethod
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
