A machine learning-based assessment strategy for Artemisia argyi Folium: Exploring the dose-effect relationship...

Artemisia argyi Folium (AAF) has been traditionally used for its medicinal properties, and understanding its anti-inflammatory activity across different regions is crucial for its effective application in herbal medicine.

This study aims to evaluate the anti-inflammatory activity of AAF sourced from various regions and to identify quality markers (Q-markers) that reflect regional variations in its chemical composition and pharmacological efficacy.

Ten chemical components of AAF from nine major production areas were analyzed by high-performance liquid chromatography (HPLC). Ethanol extracts were tested in a lipopolysaccharide-induced RAW 264.7 macrophage inflammation model to evaluate their anti-inflammatory activity through immunofluorescence, flow cytometry, and the measurement of inflammatory factors such as TNF-α, IL-1β, IL-6, and NO. PCR was used to verify the expression of these inflammatory factors, and double matrix correlation analysis linked the chemical components to their pharmacological activity. Advanced machine learning models, including Lasso, Ridge, and Elastic Net, were employed to identify key compounds that regulate inflammation.

Six compounds were screened as Q-markers. The results demonstrated significant correlations between the chemical content of AAF, its regional origin, and its anti-inflammatory efficacy. This integrated strategy provides a panel-based quality control framework for AAF.

This integrated strategy establishes a scientific foundation for the quality control and clinical application of AAF. The findings provide a systematic evaluation method for natural medicines.