Elucidating Molecular Networks Underpinning Heterogeneity in Parkinson's Disease Progression Across Clinical Manifestation Spectrum

Parkinsons disease (PD) presents with considerable clinical heterogeneity, spanning motor and non-motor symptoms with variable progression trajectories. To investigate the molecular drivers of this variability and identify therapeutic opportunities, we conducted a multi-omics, network-based analysis of the Parkinsons Progression Markers Initiative (PPMI) cohort, with independent validation in the Parkinsons Disease Biomarkers Program (PDBP) cohort. By integrating genetic and longitudinal transcriptomic data, we constructed progression-endotype networks, each capturing trait-specific molecular signatures. These networks showed significantly greater connectivity than expected by chance and converged on established PD-associated genes, including Glucosylceramidase Beta 1 (GBA1), Apolipoprotein E (APOE) haplotype, and Leucine Rich Repeat Kinase 2 (LRRK2). Using these phenotype-informed modules, we applied a network proximity approach to systematically assess 1,595 FDA-approved drugs for repurposing potential. We prioritized 25 candidates, including Zolpidem, Alprazolam, Duloxetine and Primidone. Analysis of real-world clinical data from two large research networks further revealed consistent associations between use of these drugs and reduced incidence of PD-related outcomes. Together, these findings demonstrate the utility of progression-endotype networks for capturing PD progression biology and guiding drug repurposing. This integrative framework connects molecular mechanisms with clinical impact and may inform precision therapeutic strategies for this neurodegenerative disease.