Finding the dark matter: Large language model‐based enzyme kinetic data extractor and its validation
Open accessPublished version of a preprint
Abstract
Despite the vast number of enzymatic kinetic measurements reported across decades of biochemical literature, the majority of relational enzyme kinetic data—linking amino acid sequence, substrate identity, kinetic parameters, and assay conditions—remains uncollected and inaccessible in structured form. This constitutes a significant portion of the “dark matter” of enzymology. Unlocking these hidden data through automated extraction offers an opportunity to expand enzyme dataset diversity and size, critical for building accurate, generalizable models that drive predictive enzyme engineering. To address this limitation, we built EnzyExtract, a large language model‐powered pipeline that automates the extraction, verification, and structuring of enzyme kinetics data from scientific literature. By processing 137,892 full‐text publications (PDF/XML), EnzyExtract collected more than 218,095 enzyme–substrate–kinetics entries, including 218,095 k cat and 167,794 K m values. These entries are mapped to enzymes spanning 3569 unique four‐digit EC numbers, with a total of 84,464 entries assigned at least a first‐digit EC number. EnzyExtract identified 89,544 unique kinetic entries ( k cat and K m combined) absent from BRENDA, significantly expanding the known enzymology dataset. The newly curated dataset was compiled into a database named EnzyExtractDB. EnzyExtract demonstrates high accuracy when benchmarked against manually curated datasets and strong consistency with BRENDA‐derived data. To create model‐ready datasets, enzyme and substrate sequences were aligned to UniProt and PubChem, yielding 92,286 high‐confidence, sequence‐mapped kinetic entries. To assess the practical utility of our dataset, we retrained several state‐of‐the‐art k cat predictors (including MESI, DLKcat, and TurNuP) using EnzyExtractDB. Across held‐out test sets, all models demonstrate improved predictive performance in terms of RMSE, MAE, and R 2 , highlighting the value of high‐quality, large‐scale, literature‐derived EnzyExtractDB for enhancing predictive modeling of enzyme kinetics. The EnzyExtract source code and the database are openly available at https://github.com/ChemBioHTP/EnzyExtract , and an interactive demo can be accessed via Google Colab at https://colab.research.google.com/drive/1MwKSEZzLPNOseksRshbzkkFoO_cgJhva .
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
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · see citing works
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
UniProt: a hub for protein information
The UniProt Consortium (2015) · Nucleic Acids Research · doi:10.1093/nar/gku989 · Cited by 5,106 · see citing works
Kim, S., Chen, J., Cheng, T., Gindulyte, A., He, J., He, S., Li, Q., Shoemaker, B. A., Thiessen, P. A., Yu, B., et al. (2023) · Nucleic Acids Research · doi:10.1093/nar/gkac956 · Cited by 3,443 · see citing works
Deep learning-based kcat prediction enables improved enzyme-constrained model reconstruction
Li, F., Yuan, L., Lu, H., Li, G., Chen, Y., Engqvist, M. K. M., Kerkhoven, E. J., & Nielsen, J. (2022) · Nature Catalysis · doi:10.1038/s41929-022-00798-z · Cited by 466 · see citing works
Enzyme function prediction using contrastive learning
Yu, T., Cui, H., Li, J. C., Luo, Y., Jiang, G., & Zhao, H. (2023) · Science · doi:10.1126/science.adf2465 · Cited by 451 · see citing works
SABIO-RK--database for biochemical reaction kinetics
Wittig, U., Kania, R., Golebiewski, M., Rey, M., Shi, L., Jong, L. M., Algaa, E., Weidemann, A., Sauer‐Danzwith, H., Mir, S., et al. (2011) · Nucleic Acids Research · doi:10.1093/nar/gkr1046 · Cited by 260 · see citing works
Revealing enzyme functional architecture via high-throughput microfluidic enzyme kinetics
Markin, C. J., Mokhtari, D. A., Sunden, F., Appel, M. J., Akiva, E., Longwell, S. A., Sabatti, C., Herschlag, D., & Fordyce, P. M. (2021) · Science · doi:10.1126/science.abf8761 · Cited by 222 · see citing works
BRENDA: a resource for enzyme data and metabolic information
Schomburg, I., Chang, A., Hofmann, O., Ebeling, C., Ehrentreich, F., & Schomburg, D. (2002) · Trends in Biochemical Sciences · doi:10.1016/s0968-0004(01)02027-8 · Cited by 216 · see citing works
Turnover number predictions for kinetically uncharacterized enzymes using machine and deep learning
Kroll, A., Rousset, Y., Hu, X.-P., Liebrand, N. A., & Lercher, M. J. (2023) · Nature Communications · doi:10.1038/s41467-023-39840-4 · Cited by 186 · cited 3× · see citing works
“TurNuP (Kroll et al., 2023) incorporates protein language models and attention mechanisms, splitting data by enzyme so that each unique enzyme sequence appears in only one dataset partition, with test performance further stratified by sequence identity to the training set.”
PubTables-1M: Towards comprehensive table extraction from unstructured documents
Smock, B., Pesala, R., & Abraham, R. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.00459 · Cited by 162 · see citing works
From text to insight: large language models for chemical data extraction
Schilling-Wilhelmi, M., Ríos-García, M., Shabih, S., Gil, M. V., Miret, S., Koch, C. T., Marquez, J. A., & Jablonka, K. M. (2024) · Chemical Society Reviews · doi:10.1039/d4cs00913d · Cited by 155 · see citing works
Atom-level enzyme active site scaffolding using RFdiffusion2
Ahern, W., Yim, J., Tischer, D., Salike, S., Woodbury, S. M., Kim, D., Kalvet, I., Kipnis, Y., Coventry, B., Altae-Tran, H., et al. (2025) · Nature Methods · doi:10.1038/s41592-025-02975-x · Cited by 131 · see citing works
Rajan, K., Brinkhaus, H. O., Agea, M. I., Zielesny, A., & Steinbeck, C. (2023) · Nature Communications · doi:10.1038/s41467-023-40782-0 · Cited by 79 · see citing works
STRENDA DB: enabling the validation and sharing of enzyme kinetics data
Swainston, N., Baici, A., Bakker, B. M., Cornish-Bowden, A. J., Fitzpatrick, P. F., Halling, P. J., Leyh, T. S., O’Donovan, C., Raushel, F. M., Reschel, U., et al. (2018) · FEBS Journal · doi:10.1111/febs.14427 · Cited by 63 · see citing works
Physics-based modeling in the new era of enzyme engineering
Jurich, C. P., Shao, Q., Ran, X., & Yang, Z. J. (2025) · Nature Computational Science · doi:10.1038/s43588-025-00788-8 · Cited by 47 · see citing works
Self-citationEnzyHTP: A High-Throughput Computational Platform for Enzyme Modeling
Shao, Q., Jiang, Y., & Yang, Z. J. (2022) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.1c01424 · Cited by 43 · see citing works
Self-citationEnzymeML: seamless data flow and modeling of enzymatic data
Lauterbach, S., Dienhart, H., Range, J., Malzacher, S., Spöring, J., Rother, D., Pinto, M. F., Martins, P., Lagerman, C. E., Bommarius, A., et al. (2023) · Nature Methods · doi:10.1038/s41592-022-01763-1 · Cited by 40 · see citing works
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 · see citing works
Self-citationRan, X., Jiang, Y., Shao, Q., & Yang, Z. J. (2023) · Chemical Science · doi:10.1039/d3sc02752j · Cited by 25 · see citing works
Self-citationSmith, N. S. S., Yuan, X., Melissinos, C., & Moghe, G. D. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btae756 · Cited by 24 · see citing works
EnzyHTP Computational Directed Evolution with Adaptive Resource Allocation
Shao, Q., Jiang, Y., & Yang, Z. J. (2023) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.3c00618 · Cited by 23 · see citing works
Self-citationHigh-throughput prediction of enzyme promiscuity based on substrate–product pairs
Xing, H., Cai, P., Liu, D., Han, M., Liu, J., Le, Y., Zhang, D., & Hu, Q. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae089 · Cited by 20 · see citing works
Shao, Q., Hollenbeak, A. C., Jiang, Y., Ran, X., Bachmann, B. O., & Yang, Z. J. (2025) · Chem Catalysis · doi:10.1016/j.checat.2025.101334 · Cited by 19 · see citing works
Self-citationEnzChemRED, a rich enzyme chemistry relation extraction dataset
Lai, P., Coudert, E., Aimo, L., Axelsen, K. B., Breuza, L., de Castro, E., Feuermann, M., Morgat, A., Pourcel, L., Pedruzzi, I., et al. (2024) · Scientific Data · doi:10.1038/s41597-024-03835-7 · Cited by 12 · see citing works
Turnover number predictions for kinetically uncharacterized enzymes using machine and deep learningPreprint
Kroll, A., Hu, X.-P., Liebrand, N. A., & Lercher, M. J. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.11.10.516024 · Cited by 9 · see citing works
Jiang, J., Hu, J., Xie, S., Guo, M.-H., Dong, Y., Fu, S., Jiang, X., Yue, Z., Shi, J., Zhang, X., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.03.03.641178 · Cited by 8 · see citing works
Multi-purpose enzyme-substrate interaction prediction with progressive conditional deep learningPreprint
Nie, Z., Zhang, H., Jiang, H., Liu, Y., Huang, X., Xu, F., Tian, Y., Chen, J., & Zhang, W. (2024) · Research Square · doi:10.21203/rs.3.rs-5516445/v1 · Cited by 6 · see citing works
Harnessing natural evolution and computation towards systems enzymology
Xie, W. J., & Warshel, A. (2023) · Biophysical Journal · doi:10.1016/j.bpj.2022.11.2480 · Cited by 1 · see citing works
Flagged first, then by citations
1 of the 31 references couldn’t be matched to a known paper.
Cited by
Sorted by citations
From sequence to function: Bridging single-molecule kinetics and molecular diversity
Kapanidis, A. N., Muras, L., Sreenivasa, K., Hazra, J. P., van Noort, J., Joo, C., & Deindl, S. (2026) · Science · doi:10.1126/science.adv4503 · Cited by 23 · see citing works
Long, Y., Abbasinejad, F., Li, F.-Z., Reinprecht, P., Wittmann, B. J., Kennemur, J. L., Carder, H. M., Yang, J., Lambert, T., O’Meara, R. L., et al. (2025) · Nucleic Acids Research · doi:10.1093/nar/gkaf1142 · Cited by 9 · see citing works
Zhou, R. H. F. L., Ma, H., Xin, T., Miao, Q., Zou, L., Hu, Q., Cheng, H., Guo, J., Mu, Y., Wang, S., et al. (2026) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6c00090 · Cited by 3 · see citing works
Kell, D. B., & Roberts, I. (2026) · Chemical Society Reviews · doi:10.1039/d5cs01387a
Liu, H. (2026) · Computational Biology and Chemistry · doi:10.1016/j.compbiolchem.2026.109222
Irons, J. L., Cooper, P. S., Bölücü, N., Duenser, A., Timmer, R. C., Yang, H., Lee, C., Jin, B., & Wan, S. (2026) · ACM SIGIR International Conference on Theory of Information Retrieval (ICTIR) · doi:10.1145/3805713.3820430
Li, G., Huang, J., Gao, X., Zhou, Y., & Zhu, L. (2026) · ENGINEERING Energy · doi:10.1007/s11708-026-1095-8
Zhai, D., Wang, S., Li, Y., Yuan, Q., PengYang, D., & Liu, A. (2026) · Advances in biochemical engineering, biotechnology · doi:10.1007/10_2026_347
François, J., Moor, B. D., & van Noort, V. (2026) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6c02109
Harnessing artificial intelligence to understand enzyme catalysis and discover new enzymes
Brownless, A.-L. R., & Kamerlin, S. C. L. (2026) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2026.103422
- Authors
- Wei, G., Ran, X., AI‐Abssi, R., & Yang, Z.
- Affiliations
- Vanderbilt University (US)
- Department of Chemistry Vanderbilt University Nashville Tennessee USA
- Corresponding author
- Yang, Zhongyue
- Venue
- Protein Science
- Publisher
- Wiley
- Published
- 2025-08-15
- Volume
- 34
- Issue
- 9
- Pages
- e70251-e70251
- ISSN
- 0961-8368 (ISSN-L)
- 1469-896X
- PMC
- PMC12355964
- OpenAlex
- W4413137565
- Cited by
- 10
- References
- 31
- Citations by year
- 2026: 9 · 2025: 1
- FWCI
- 3.01
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Protein Science (publishedVersion, cc-by)
- ChemRxiv (acceptedVersion, cc-by-nc-nd)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Funders
- Arnold and Mabel Beckman Foundation
- National Science Foundation (BIO200057.102)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Biomedical Text Mining and Ontologies
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Enzyme Catalysis and Immobilization
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
- Keywords
- enzyme kinetics, large language models, BRENDA, Michaelis constant, literature curation, data curation
- MeSH terms
- Large Language Models, Enzymes, Kinetics, Software, Databases, Protein
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
