Related Experiment Video
Updated: Jan 13, 2026
07:46
On-site DNA Detection of Trypanosomatid Parasites and Nosema ceranae Through Alkaline Lysis Coupled to RPA/CRISPR/Cas12a System
Published on: July 18, 2025
832
Precise Discrimination Between Rape Honey and Acacia Honey Based on Sugar and Amino Acid Profiles Combined with
Chenyu Sun1, Fei Pan2, Wenli Tian2
1College of Science, China Agricultural University, Beijing 100083, China.
Foods (Basel, Switzerland)
|January 10, 2026
Summary
Accurate honey authentication is crucial. This study developed a machine learning method to precisely distinguish acacia honey from adulterated rape honey using chemical profiles, achieving high prediction accuracy.
Area of Science:
- Food Chemistry
- Analytical Chemistry
- Computational Biology
Background:
- Honey adulteration, particularly of high-value unifloral honeys like acacia, poses significant market and consumer risks.
- Distinguishing visually similar honeys such as acacia and rape honey is challenging due to overlapping chemical characteristics.
Purpose of the Study:
- To develop a precise method for discriminating between rape honey and acacia honey.
- To leverage chemical profiling and machine learning for accurate honey authentication.
Main Methods:
- Collected 542 honey samples from China.
- Quantified 12 sugars and 20 amino acids using Ultra-Performance Liquid Chromatography-Tandem Mass Spectrometry (UPLC-MS/MS).
- Applied six machine learning algorithms with 10-fold cross-validation and ADASYN oversampling for classification.
Main Results:
- Multivariate analysis showed significant compositional differences between rape and acacia honey, despite some overlap.
- Machine learning models achieved high prediction accuracies (98% for rape honey, 100% for acacia honey) on an independent test set.
- SHAP analysis identified key discriminators: fructose, turanose, glucose, and Gamma-Aminobutyric Acid (GABA).
Conclusions:
- The developed chemical profiling and machine learning framework enables precise discrimination between acacia and rape honey.
- The study provides a robust technical foundation for honey quality control and anti-fraud initiatives.
- A user-friendly web application was created for rapid, on-site honey authentication.
