The DOI matches its Crossref record; 26 of the 26 references it cites were found in a registry or an index.
Ankh ☥: Optimized Protein Language Model Unlocks General-Purpose Modelling
Open accessAbstract
As opposed to scaling-up protein language models (PLMs), we seek improving performance via protein-specific optimization. Although the proportionality between the language model size and the richness of its learned representations is validated, we prioritize accessibility and pursue a path of data-efficient, cost-reduced, and knowledge-guided optimization. Through over twenty experiments ranging from masking, architecture, and pre-training data, we derive insights from protein-specific experimentation into building a model that interprets the language of life, optimally. We present Ankh, the first general-purpose PLM trained on Google’s TPU-v4 surpassing the state-of-the-art performance with fewer parameters (<10% for pre-training, <7% for inference, and <30% for the embedding dimension). We provide a representative range of structure and function benchmarks where Ankh excels. We further provide a protein variant generation analysis on High-N and One-N input data scales where Ankh succeeds in learning protein evolutionary conservation-mutation trends and introducing functional diversity while retaining key structural-functional characteristics. We dedicate our work to promoting accessibility to research innovation via attainable resources.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2019) · arXiv · doi:10.48550/arxiv.1910.10683 · Cited by 10,929
ColabFold - Making protein folding accessible to all
Mirdita, M., Schütze, K., Moriwaki, Y., Heo, L., Ovchinnikov, S., & Steinegger, M. (2021) · doi:10.1101/2021.08.15.456425 · Cited by 9,391
Biopython: freely available python tools for computational molecular biology and bioinformatics
Cock, P. J. A., Antao, T., Chang, J. T., Chapman, B. A., Cox, C. J., Dalke, A., Friedberg, I., Hamelryck, T., Kauff, F., Wilczynski, B., et al. (2009) · Bioinformatics · doi:10.1093/bioinformatics/btp163 · Cited by 6,080 · cited 2×
“We compute the identity via BioPython’s pairwise2 module where we obtain the global alignment between the generated variants (gen) and the original ones (nat) [13].”
MethodALBERT: A Lite BERT for Self-supervised Learning of Language Representations
Lan, Z., Chen, M., Goodman, S., Gimpel, K., Sharma, P., & Soricut, R. (2019) · arXiv · doi:10.48550/arxiv.1909.11942 · Cited by 4,308 · cited 2×
“, BERT [21], ALBERT [22], and Electra [23]) and decoder-only transformers (e.”
BackgroundRives, A., Meier, J., Sercu, T., Goyal, S., Lin, Z., Liu, J., Guo, D., Ott, M., Zitnick, C. L., Ma, J., et al. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2016239118 · Cited by 3,286
Rives, A., Meier, J., Sercu, T., Goyal, S., Lin, Z., Liu, J., Guo, D., Ott, M., Zitnick, C. L., Ma, J., et al. (2019) · doi:10.1101/622803 · Cited by 3,280 · cited 1×
“Notably, PLM sizes have jumped from ∼ 10(6) [4] to ∼ 10(9) [10] parameters recently.”
BackgroundRoFormer: Enhanced transformer with rotary position embedding
Su, J., Ahmed, M., Lu, Y., Pan, S., Bo, W., & Liu, Y. (2024) · Neurocomputing · doi:10.1016/j.neucom.2023.127063 · Cited by 1,957
RoFormer: Enhanced Transformer with Rotary Position Embedding
Su, J., Lu, Y., Pan, S., Murtadha, A., Wen, B., & Liu, Y. (2021) · arXiv · doi:10.48550/arxiv.2104.09864 · Cited by 1,952 · cited 1×
“Using a fixed positional embedding has several limitations such as not being able to extract embeddings for tokens exceeding the maximum length of the pre-defined positional embedding dimension, correlated to the embedding dimension [59].”
BackgroundUniRef clusters: a comprehensive and scalable alternative for improving sequence similarity searches
Suzek, B. E., Wang, Y., Huang, H., McGarvey, P. B., Wu, C. H., & the UniProt Consortium (2015) · Bioinformatics · doi:10.1093/bioinformatics/btu739 · Cited by 1,928
ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators
Clark, K., Luong, M.-T., Le, Q. V., & Manning, C. D. (2020) · arXiv · doi:10.48550/arxiv.2003.10555 · Cited by 1,844
Unified 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
Predicting Parameters in Deep Learning
Denil, M., Shakibi, B., Dinh, L., Ranzato, M., & de Freitas, N. (2013) · arXiv · doi:10.48550/arxiv.1306.0543 · Cited by 853
Evaluating 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 · cited 4×
“(i) In GB1 Fitness Prediction, the input is a protein sequence corresponding to mutations at four possible positions and the output is a regression score indicating the fitness prediction 4.2.1 Tasks and Datasets The emergence of transformers as powerful self-surprised models for proteins has motivated significant eff…”
BackgroundMethodInfluentialModeling 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
Self-citationModeling aspects of the language of life through transfer-learning protein sequences
Heinzinger, M., Elnaggar, A., Wang, Y., Dallago, C., Nechaev, D., Matthes, F., & Rost, B. (2019) · BMC Bioinformatics · doi:10.1186/s12859-019-3220-8 · Cited by 606
Self-citationLocal fitness landscape of the green fluorescent protein
Sarkisyan, K. S., Bolotin, D. A., Meer, M. V., Usmanova, D. R., Mishin, A. S., Sharonov, G. V., Ivankov, D. N., Bozhanova, N. G., Baranov, M. S., Soylemez, O., et al. (2016) · Nature · doi:10.1038/nature17995 · Cited by 591 · cited 2×
“Fluorescence Prediction (FluP): This regression task evaluates the fluorescence intensity of green fluorescent protein mutants annotated by Sarkisyan et al [28].”
BackgroundCATH: increased structural coverage of functional space
Sillitoe, I., Bordin, N., Dawson, N., Waman, V. P., Ashford, P., Scholes, H. M., Pang, C. S. M., Woodridge, L., Rauer, C., Sen, N., et al. (2021) · Nucleic Acids Research · doi:10.1093/nar/gkaa1079 · Cited by 509
Critical assessment of methods of protein structure prediction (CASP)—Round XIII
Kryshtafovych, A., Schwede, T., Topf, M., Fidelis, K., & Moult, J. (2019) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.25823 · Cited by 506
Critical assessment of methods of protein structure prediction (CASP)—Round XIV
Kryshtafovych, A., Schwede, T., Topf, M., Fidelis, K., & Moult, J. (2021) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.26237 · Cited by 503
ProGen2: Exploring the boundaries of protein language models
Nijkamp, E., Ruffolo, J. A., Weinstein, E. N., Naik, N., & Madani, A. (2023) · Cell Systems · doi:10.1016/j.cels.2023.10.002 · Cited by 480
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 3×
“The analysis promoted utilizing UniRef50 [17] over UniRef100 [17] and BFD [3, 18].”
MethodInfluentialExpanding functional protein sequence spaces using generative adversarial networks
Repecka, D., Jauniskis, V., Karpus, L., Rembeza, E., Rokaitis, I., Zrimec, J., Poviloniene, S., Laurynenas, A., Viknander, S., Abuajwa, W., et al. (2021) · Nature Machine Intelligence · doi:10.1038/s42256-021-00310-5 · Cited by 388
CoV-AbDab: the coronavirus antibody database
Raybould, M. I. J., Kovaltsuk, A., Marks, C., & Deane, C. M. (2021) · Bioinformatics · doi:10.1093/bioinformatics/btaa739 · Cited by 383
Transformer protein language models are unsupervised structure learners
Rao, R., Meier, J., Sercu, T., Ovchinnikov, S., & Rives, A. (2020) · doi:10.1101/2020.12.15.422761 · Cited by 341
BERTology Meets Biology: Interpreting Attention in Protein Language Models
Vig, J., Madani, A., Varshney, L. R., Xiong, C., Socher, R., & Rajani, N. F. (2020) · doi:10.1101/2020.06.26.174417 · Cited by 338
Cited by
Bilingual language model for protein sequence and structure
Heinzinger, M., Weissenow, K., Sanchez, J. G., Henkel, A., Mirdita, M., Steinegger, M., & Rost, B. (2024) · NAR Genomics and Bioinformatics · doi:10.1093/nargab/lqae150 · Cited by 311
Self-citationMachine learning for functional protein design
Notin, P., Rollins, N., Gal, Y., Sander, C., & Marks, D. (2024) · Nature Biotechnology · doi:10.1038/s41587-024-02127-0 · Cited by 270
Fine-tuning protein language models boosts predictions across diverse tasks
Schmirler, R., Heinzinger, M., & Rost, B. (2024) · Nature Communications · doi:10.1038/s41467-024-51844-2 · Cited by 229
“Other correlation coefficients are not beneficial (as ranking is the most interesting here) and error metrics like MSE are not informative for arbitrary fitness scores.”
Self-citationConvolutions are competitive with transformers for protein sequence pretraining
Yang, K. K., Fusi, N., & Lu, A. X. (2024) · Cell Systems · doi:10.1016/j.cels.2024.01.008 · Cited by 137
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
Metadata
- Authors
- Elnaggar, A., Essam, H., Salah-Eldin, W., Moustafa, W., Elkerdawy, M., Rochereau, C., & Rost, B.
- Venue
- bioRxiv (Cold Spring Harbor Laboratory)
- Publisher
- openRxiv
- Published
- 2023-01-18
- OpenAlex
- W4317374308
- Cited by
- 184
- References
- 26
- Type
- posted content
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.