Finding the dark matter: Large language model‐based enzyme kinetic data extractor and its validation
Protein Science · 2025-09 · 34 (9)
DOI: 10.1002/pro.70251 ↗Kurz gefasst
Fachartikel in der Zeitschrift Protein Science, online erschienen am 15.08.2025. Verfasst von Galen Wei und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Chemistry, Vanderbilt University, Nashville, Tennessee, USA. Hauptthemen laut MEDLINE: Enzymes, Databases, Protein, Software. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 32 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Institute of General Medical Sciences of the National Institutes of Health, NIGMS NIH HHS, Arnold and Mabel Beckman Foundation.
Abstract (Originalsprache)
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 .
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Chemistry, Vanderbilt University, Nashville, Tennessee, USA.
- Center for Structural Biology, Vanderbilt University, Nashville, Tennessee, USA.
- Vanderbilt Institute of Chemical Biology, Vanderbilt University, Nashville, Tennessee, USA.
- Data Science Institute, Vanderbilt University, Nashville, Tennessee, USA.
- Department of Chemical and Biomolecular Engineering, Vanderbilt University, Nashville, Tennessee, USA.
ORCID-Profile: Zhongyue Yang
Förderung
- National Institute of General Medical Sciences of the National Institutes of Health
- NIGMS NIH HHS
- Arnold and Mabel Beckman Foundation
- OpenAI Researcher Access Program
- National Science Foundation
Zitieren
APA
Wei, G., Ran, X., Ai-Abssi, R., & Yang, Z. (2025). Finding the dark matter: Large language model‐based enzyme kinetic data extractor and its validation. Protein Science, 34(9). https://doi.org/10.1002/pro.70251
Vancouver
Wei G, Ran X, Ai-Abssi R, Yang Z. Finding the dark matter: Large language model‐based enzyme kinetic data extractor and its validation. Protein Science. 2025;34(9). doi:10.1002/pro.70251
BibTeX
@article{Wei2025,
author = {Wei, Galen and Ran, Xinchun and Ai-Abssi, Runeem and Yang, Zhongyue},
title = {Finding the dark matter: Large language model‐based enzyme kinetic data extractor and its validation},
journal = {Protein Science},
year = {2025},
volume = {34},
number = {9},
doi = {10.1002/pro.70251}
}
EUVIMED