AI-assisted integrative framework combining microarray data analysis and cerebrospinal fluid pharmacology for...

Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by a gradual decline in cognitive function, with a complex pathogenesis involving multiple targets and signaling pathways, and current therapeutic strategies remain insufficient to achieve satisfactory outcomes. Psoraleae Fructus (PF) has been reported to exhibit potential therapeutic effects against AD. However, its multi-target mechanisms of action have not yet been systematically elucidated. In this study, an AI-assisted integrative computational and experimental framework was established to comprehensively investigate the molecular basis of PF intervention in AD. Targeted constituents of PF were first identified using UPLC-Q Exactive Orbitrap high-resolution mass spectrometry, followed by the integration of Artificial intelligence modeling, microarray data, and network pharmacology to screen hub targets. Cerebrospinal fluid-based pharmacological assays, together with molecular docking, molecular dynamics simulations, microarray validation, and western blot, were subsequently employed to validate. The results demonstrated that PF markedly modulated the expression of GSK3β and PPARγ, thereby regulating core AD-related pathological markers, including p-tau, Aβ₄₂, β-secretase, and inflammatory mediators. Collectively, this study delineated a multi-target regulatory network underlying the anti-AD effects of PF and provided a robust theoretical foundation for its further translational investigation.