Optimization of the Extraction Process of Effective Components of Eleutherococcus senticosus Using Mathematical Models

• Published in: Journal of food quality 2024-09, Vol.2024 (1)
• Main Authors: Zhang, Shunjie, Wang, Yudan, Liu, Tianyang, Li, Xinzhi, Wang, Meiliang, Qiu, Youwen
• Format: Article
• Language: English
• Subjects: Algorithms; Analysis; Carbon dioxide; Chromatography; Classification; Clustering; Data analysis; Decision trees; Experimental design; High performance liquid chromatography; Liquid chromatography; Machine learning; Mathematical models; Medicinal plants; Medicine, Botanic; Medicine, Herbal; Methods; Prediction models; Variables
• ISSN: 0146-9428; 1745-4557
• DOI: 10.1155/2024/5538373

Eleutheroside B, eleutheroside E, and isofraxidin are all important active ingredients for medicinal purposes of Eleutherococcus senticosus (Rupr. and Maxim).To improve the efficacy and quality of Chinese medicinal materials, optimizing the extraction process is necessary. By improving and innovating the process, the utilization efficiency of medicinal materials can be enhanced, promoting the effective utilization of Chinese medicinal materials. This experiment focused on three traditional components of Eleutherococcus senticosus, a medicinal plant known for its various bioactive properties. High-performance liquid chromatography and response surface methodology were employed to investigate the effects of extraction parameters on the yield of these components. Mathematical model analysis was used to optimise the response surface experimental design and to determine the exact extraction conditions for each ingredient. In the extraction method prediction model, the method described in the best group extracted more of the three active ingredients. The extraction rate of the components was increased by 1.33 percent compared to the conventional extraction method. Meanwhile, the R-square of this prediction model was as high as 0.938. A validated method was used to quantify the three components in 21 samples of Eleutherococcus senticosus from different origins. The results revealed significant variations in the component content among the samples. The recall F1 of cross-validation sets in the GA-decision tree of the classifier reaches 0.833. K-means cluster analysis was performed to classify the samples based on their component profile, providing a basis for quality assessment. The novel content evaluation model developed in this study enabled a reliable and visual comparison of the component differences among the samples and extraction conditions, determining the optimal extraction parameters for each component of Eleutherococcus senticosus. This study offers new techniques and insights for the development and utilization of natural active ingredients from Chinese herbs, with important scientific and practical implications.