To combat fraudulent sales of low-aged ginseng disguised as high-aged ginseng, scientists from Shanghai University of Traditional Chinese Medicine in China, developed machine learning (ML) models to predict the ages of ginseng samples. Their work was published in the Journal of Separation Science (1).
Fresh Korean Ginseng over wooden plate | Image Credit: © leungchopan - stock.adobe.com
The study aims to differentiate mountain-cultivated ginseng by age. Ginseng has been studied for its multiple health benefits, including boosting the immune system and fighting conditions like colds, flu, and cancer. Mountain-cultivated ginseng is usually harvested after 10 years and can produce more berries and seeds than wild ginseng plants, or ginseng plants harvested after 15 years. This difference has led to issues in the ginseng market, with some fraudulently selling low-aged cultivated ginseng and disguising it as high-aged.
For this experiment, 98 ginseng samples were analyzed using liquid chromatography–mass spectrometry (LC–MS), with multivariate statistical analysis used to identify patterns between samples and influential components. ML models were also created to assist in this process. Untargeted metabolomic analysis divided samples that were between 4–20 years into three age groups, with 22 age-dependent biomarkers discovered to differentiate between said age groups.
From there, three more ML models were made to predict new samples, which eventually led to an optimal model being selected. According to the scientists, “Some biomarkers could determine age phases according to the differentiation of mountain-cultivated ginseng samples” (1). The biomarkers were later analyzed for potential variation trends.
(1) Chen, G.; Zhang, H.; Jiang, J.; Chen, S.; Zhang, H.; Zhang, G.; Zheng, C.; Xu, H. Metabolomics approach to growth-age discrimination in mountain-cultivated ginseng (Panax ginseng C. A. Meyer) using ultra-high-performance liquid chromatography coupled with quadrupole-time-of-flight mass spectrometry. J. Sep. Sci. 2023, 2300445. DOI: https://doi.org/10.1002/jssc.202300445
(2) American ginseng. Icahn School of Medicine at Mount Sinai 2023. https://www.mountsinai.org/health-library/herb/american-ginseng (accessed 2023-11-09)
(3) Differences between Wild and Cultivated Ginseng. National Ginseng LLC 2023. https://www.wildgrown.com/index.php/general-information-about-ginseng/107-differences-between-wild-and-cultivated-ginseng (accessed 2023-11-09)
Rethinking Chromatography Workflows with AI and Machine Learning
April 1st 2025Interest in applying artificial intelligence (AI) and machine learning (ML) to chromatography is greater than ever. In this article, we discuss data-related barriers to accomplishing this goal and how rethinking chromatography data systems can overcome them.
Advances in Non-Targeted Analysis for PFAS in Environmental Matrices
March 27th 2025David Megson from Manchester Metropolitan University in Manchester, UK, spoke to LCGC International about the latest developments in non-targeted analysis (NTA) of per- and polyfluoroalkyl substances (PFAS) in environmental matrices based on a recent systematic review paper he has collaboratively published (1).
Study Explores Thin-Film Extraction of Biogenic Amines via HPLC-MS/MS
March 27th 2025Scientists from Tabriz University and the University of Tabriz explored cellulose acetate-UiO-66-COOH as an affordable coating sorbent for thin film extraction of biogenic amines from cheese and alcohol-free beverages using HPLC-MS/MS.
Quantifying Microplastics in Meconium Samples Using Pyrolysis–GC-MS
March 26th 2025Using pyrolysis-gas chromatography and mass spectrometry, scientists from Fudan University and the Putuo District Center for Disease Control and Prevention detected and quantified microplastics in newborn stool samples.