ProteinGLUE multi-task benchmark suite for self-supervised protein modeling
Open accessPublished version of a preprint
Abstract
Self-supervised language modeling is a rapidly developing approach for the analysis of protein sequence data. However, work in this area is heterogeneous and diverse, making comparison of models and methods difficult. Moreover, models are often evaluated only on one or two downstream tasks, making it unclear whether the models capture generally useful properties. We introduce the ProteinGLUE benchmark for the evaluation of protein representations: a set of seven per-amino-acid tasks for evaluating learned protein representations. We also offer reference code, and we provide two baseline models with hyperparameters specifically trained for these benchmarks. Pre-training was done on two tasks, masked symbol prediction and next sentence prediction. We show that pre-training yields higher performance on a variety of downstream tasks such as secondary structure and protein interaction interface prediction, compared to no pre-training. However, the larger model does not outperform the smaller model. We expect the ProteinGLUE benchmark dataset introduced here, together with the two baseline pre-trained models and their performance evaluations, to be of great value to the field of protein sequence-based property prediction. Availability: code and datasets from https://github.com/ibivu/protein-glue .
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
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.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764 · see citing works
Deep Contextualized Word Representations
Peters, M., Neumann, M., Iyyer, M., Gardner, M., Clark, C., Lee, K., & Zettlemoyer, L. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers) · doi:10.18653/v1/n18-1202 · Cited by 11,983 · see citing works
The Power of Scale for Parameter-Efficient Prompt Tuning
Lester, B., Al-Rfou, R., & Constant, N. (2021) · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2021.emnlp-main.243 · Cited by 3,096 · see citing works
The Power of Scale for Parameter-Efficient Prompt Tuning
Lester, B., Al-Rfou, R., & Constant, N. (2021) · arXiv · doi:10.48550/arxiv.2104.08691 · Cited by 3,092
Highly accurate protein structure prediction for the human proteome
Tunyasuvunakool, K., Adler, J., Wu, Z., Green, T., Zielinski, M., Žídek, A., Bridgland, A., Cowie, A., Meyer, C., Laydon, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03828-1 · Cited by 3,025 · cited 1× · see citing works
“Notwithstanding the enormous progress that AlphaFold2 represents, reliable structural information remains unavailable for many proteins 56,57 .”
BackgroundProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning
Elnaggar, A., Heinzinger, M., Dallago, C., Rehawi, G., Wang, Y., Jones, L., Gibbs, T., Feher, T., Angerer, C., Steinegger, M., et al. (2022) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2021.3095381 · Cited by 2,329 · cited 7× · see citing works
“Although the CB513 dataset 55 has been called redundant and outdated 11 , it is commonly used to evaluate model quality on the prediction of secondary structure in eight classes as it is sufficiently difficult to avoid ceiling effects.”
BackgroundMethodInfluentialUnified rational protein engineering with sequence-based deep representation learning
Alley, E. C., Khimulya, G., Biswas, S., AlQuraishi, M., & Church, G. M. (2019) · Nature Methods · doi:10.1038/s41592-019-0598-1 · Cited by 1,131 · cited 5× · see citing works
“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 .”
BackgroundMethodInfluentialEvaluating Protein Transfer Learning with TAPE
Rao, R., Bhattacharya, N., Thomas, N., Duan, Y., Chen, X., Canny, J., Abbeel, P., & Song, Y. S. (2019) · doi:10.1101/676825 · Cited by 719 · see citing works
Modeling the language of life – Deep Learning Protein Sequences
Heinzinger, M., Elnaggar, A., Wang, Y., Dallago, C., Nechaev, D., Matthes, F., & Rost, B. (2019) · doi:10.1101/614313 · Cited by 606 · see citing works
Protein-level assembly increases protein sequence recovery from metagenomic samples manyfold
Steinegger, M., Mirdita, M., & Söding, J. (2019) · Nature Methods · doi:10.1038/s41592-019-0437-4 · Cited by 445 · cited 1× · see citing works
“…sequences 42 , UniRep is trained on 24 million UniRef50 sequences, ESM-1b is pre-trained on 250 million UniParc sequences, and Elnaggar et al. 11 even train their models on the BFD dataset containing over 2 billion sequences 43,44 , and on the UniRef100 database with 216 million sequences.”
MethodRao, R., Liu, J., Verkuil, R., Meier, J., Canny, J. F., Abbeel, P., Sercu, T., & Rives, A. (2021) · doi:10.1101/2021.02.12.430858 · Cited by 350 · see citing works
Learning protein sequence embeddings using information from structure
Bepler, T., & Berger, B. (2019) · arXiv · doi:10.48550/arxiv.1902.08661 · Cited by 219 · see citing works
Flagged first, then by citations
- Authors
- Capel, H., Weiler, R., Dijkstra, M., Vleugels, R., Bloem, P., & Feenstra, K. A.
- Affiliations
- Vrije Universiteit Amsterdam (NL)
- Corresponding author
- Feenstra, K. Anton
- Venue
- Scientific Reports
- Published
- 2022-09-26
- Volume
- 12
- Issue
- 1
- Pages
- 16047-16047
- ISSN
- 2045-2322 (ISSN-L)
- PMC
- PMC9512797
- OpenAlex
- W4297243376
- Cited by
- 11
- References
- 12
- Citations by year
- 2026: 1 · 2025: 3 · 2024: 3 · 2023: 2 · 2022: 2
- FWCI
- 0.82
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- Scientific Reports (publishedVersion, cc-by)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc)
- Research Square (acceptedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- VU Research Portal (publishedVersion, other-oa)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Data Archiving and Networked Services (DANS) (submittedVersion, other-oa)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Computational Drug Discovery Methods
- Keywords
- protein representation, benchmark suite, next sentence prediction, pre-training, secondary structure prediction, multi-task learning
- MeSH terms
- Amino Acid Sequence, Amino Acids, Natural Language Processing, Proteins
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
