Polypharmacology in Alzheimer’s Disease: Integrating AI, Network Pharmacology, and Experimental Validation

Abstract

Alzheimer’s disease (AD), the leading cause of dementia, is a progressive neurodegenerative disorder characterized by amyloid-β plaques, tau tangles, oxidative stress, mitochondrial dysfunction, and neuroinflammation. Current monotherapies, designed using a “one drug-one target” approach, provide only symptomatic relief without altering disease progression. Polypharmacology, through the development of multi-target-directed ligands (MTDLs), offers a promising therapeutic strategy that simultaneously modulates multiple interconnected pathways in AD pathology. This review highlights the integration of computational and experimental approaches in the discovery of MTDLs. In silico tools, such as molecular docking, pharmacophore modeling, QSAR, network pharmacology, and artificial intelligence, facilitate the prediction and design of polypharmacological agents. In vitro, in vivo, and omics-based studies have validated their therapeutic relevance. Hybrid molecules, including ladostigil, M-30, ASS234, donecopride, and chalcone–rivastigmine hybrids, demonstrate multi-faceted neuroprotection by targeting cholinergic dysfunction, amyloid aggregation, oxidative stress, and neuroinflammation. Despite challenges such as off-target effects and translational limitations, advances in AI-driven platforms, systems biology, and human-relevant models, such as brain organoids, are expected to accelerate the development of disease-modifying therapies. Polypharmacology represents a paradigm shift in AD treatment, moving beyond symptomatic relief towards mechanism-informed interventions with the potential to slow or halt disease progression.