Integrating network pharmacology, machine learning, and molecular docking to explore the therapeutic mechanisms of Huangqin in atopic dermatitis: A STROBE-compliant observational study

Abstract

Atopic dermatitis (AD) is a chronic inflammatory skin disorder with complex pathogenesis, and current therapies face limitations in efficacy and safety. Huangqin (Scutellaria baicalensis) exhibits anti-inflammatory properties, yet its multi-target mechanisms against AD remain unclear. A systems pharmacology approach integrating multi-omics profiling was utilized to decode Huangqin anti-AD mechanisms. First, bioactive compounds and their potential targets were systematically identified, followed by constructing compound–target networks and enriching key pathways. Then, machine learning algorithms (Support Vector Machine/Recursive Feature/Least Absolute Shrinkage and Selection Operator) were applied to prioritize hub targets from network-derived candidates. Finally, molecular docking was conducted to validate ligand–receptor binding affinity. Twenty-nine bioactive compounds were identified, interacting with 55 AD-related targets. AKT1 emerged as the most central hub in the protein–protein interaction network. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analysis revealed Huangqin potential roles in modulating bacterial infection responses and regulating pathways such as IL-17, TNF, HIF-1α, and PI3K–AKT signaling. Machine learning algorithms were applied to prioritize key genes, which highlighted AKT1, solute carrier family 6 member 4, and chemokine ligand 2 as core targets, with molecular docking confirming strong binding between wogonin, baicalein, beta-sitosterol, and these targets. These findings suggest that Huangqin exerts multi-target effects on AD, centered on AKT1-mediated signaling crosstalk, to regulate inflammatory and immune pathways. This mechanistic insight establishes a foundation for clinical translation and AKT1-focused drug development.