Large-Scale Analysis of Plasma Proteomics Reveals the Identity of Alzheimer’s and Parkinson’s Molecular Subtypes

Background: Neurodegenerative diseases such as Alzheimer’s disease (AD) and Parkinson’s disease (PD) are increasingly recognized as biologically heterogeneous disorders. Although biomarker-based frameworks have improved disease stratification, they may not fully capture the molecular diversity underlying disease mechanisms. This study aimed to define biologically meaningful AD and PD subtypes by employing large-scale plasma proteomics to characterize mechanism-driven patient stratification. Methods: We analyzed plasma proteomic data from the Global Neurological Proteomics Consortium (GNPC) dataset and performed disease-specific unsupervised clustering to identify molecular subtypes. We performed differential protein expression and pathway enrichment analyses to characterize subtype-specific and shared biological signatures. Results: Clustering analysis revealed three molecular AD subtypes and two PD subtypes, each exhibiting unique proteomic signatures despite similar demographic and cognitive profiles. AD subtypes differed in the relative involvement of immune activation, mitochondrial dysfunction, synaptic signaling, and calcium-related pathways, whereas PD subtypes showed divergence in neuroinflammatory, oxidative stress, and mitochondrial programs. Conclusions: We demonstrated stratification of AD and PD into molecular subtypes, potentially supporting scalable, mechanism-directed stratification and the development of future targeted, disease-modifying therapies.