A comprehensive strategy for prediction and quality evaluation of standardized planting herbs based on plant metabolomics coupled with extreme learning machine: Astragali Radix as an example.

Abstract

Standardizing the planting process is an effective way to control the quality stability of herbal resources, which are susceptible to external environmental factors (e.g., moisture, soil, etc.). However, how to scientifically and comprehensively assess the effects of standardized planting on plant quality and quickly test unknown samples has not been addressed. The aim of this study was to determine and compare the metabolite levels of herbs before and after standardized planting, to quickly distinguish their sources, and to evaluate their quality, using the typical herb Astragali Radix (AR) as an example. In this study, an efficient strategy using liquid chromatography-mass spectrometry (LC-MS) based on plant metabolomics combined with extreme learning machine (ELM) has been developed to efficiently distinguish and predict AR after standardized planting. Moreover, a comprehensive multi-index scoring method has been developed for the comprehensive evaluation of the quality of AR. The results confirmed that AR after standardized planting was significantly differentiated, with a relatively stable content of 43 differential metabolites, mainly including flavonoids. An ELM model was established based on LC-MS data, and the accuracy in predicting unknown samples could reach more than 90%. As expected, higher total scores were obtained for AR after standardized planting, indicating much better quality. A dual system for evaluating the impact of standardized planting on the quality of plant resources has been established, which will significantly contribute to innovation in the quality evaluation of medicinal herbs and support the selection of optimal planting conditions.