Related Experiment Video
Updated: Jul 3, 2026
14:18
A Strategy for Sensitive, Large Scale Quantitative Metabolomics
Published on: May 27, 2014
Large Language Model-Generated Dietary Metabolite Biomarker Database Drives Deep Annotation of the Human Diet
Zijun Nie1, Fujian Zheng1, Dejun Hu1
1State Key Laboratory of Natural Medicines, National R&D Center for Chinese Herbal Medicine Processing, School of Engineering, China Pharmaceutical University, Nanjing 210009, China.
Analytical Chemistry
|July 1, 2026
Summary
This study introduces a new method using large language models (LLMs) to build a dietary biomarker database, improving the accuracy and coverage of nutritional epidemiology research.
Area of Science:
- Metabolomics
- Nutritional Epidemiology
- Bioinformatics
Background:
- Accurate identification of dietary biomarkers is essential for nutritional epidemiology.
- Current public databases for LC-HRMS lack specificity, limiting biomarker annotation coverage and accuracy.
Purpose of the Study:
- To develop a novel database construction strategy and a dual-annotation workflow for enhanced dietary biomarker identification.
- To improve the accuracy and coverage of dietary metabolomics analysis.
Main Methods:
- Utilized a large language model (LLM)-based text-mining pipeline to construct the Dietary Metabolite Biomarker Database (DMBDB) from scientific literature.
- Developed a dual-annotation workflow combining a specialized LC-MS database and a structure-guided molecular networking strategy (SGMNS).
Main Results:
- The LLM pipeline successfully created the DMBDB with 4983 biomarkers and achieved a high F1 score (0.9269) for biomarker recognition.
- The dual-annotation framework identified 566 metabolites using the LC-MS database and expanded annotations to 2078 with SGMNS integration.
Conclusions:
- The LLM-driven database construction and dual-annotation workflow offer a powerful approach for high-coverage, high-accuracy dietary metabolomics.
- This strategy addresses limitations in existing databases, advancing nutritional epidemiology research.
