Multi-component analyses of raspberry: Optimization of extraction procedure and network pharmacology

Abstract

The contents of ellagic acid and kaempferol-3-O-rutinoside, the chief active components of raspberry, are considered the quality control indices of raspberry. This work employed the ant colony neural network (ACO-BPNN) to optimize their extraction processes, and the combination of network pharmacology and molecular docking technology to unveil the potential pharmacological effects of these components. Based on the single-factor test (ultrasonic time, ethanol concentration, ultrasonic temperature, and solid-liquid ratio), a factorial experiment with 4-factors and 3-levels was conducted in parallel for 3 times. The multi-factor analysis of variance results revealed high-order interactions among the factors. Then, the ACO-BPNN model was established to characterize the complex relationship of experimental data. After further verification, relative errors were all less than 8 %, implying the model's effectiveness and reliability. Moreover, with the network pharmacology, 66 key targets were screened out and mainly concentrated in PI3K-AKT, MAPK, and Ras signal pathways. Molecular docking revealed the binding sites between active components and key targets.