A Patient-Specific Metabolic Modeling Framework of Whole-Blood Transcriptomes for Early Screening of Prodromal Parkinson’s Disease

Abstract

Abstract Parkinson’s disease (PD) is a progressive neurodegenerative disorder associated with widespread metabolic dysfunction, particularly involving mitochondrial pathways. Because clinical diagnosis typically occurs only after substantial neuronal loss has already occurred, there is increasing interest in identifying molecular alterations associated with the prodromal phase of disease progression. Genome-scale metabolic models (GEMs) provide a systems-level framework for integrating transcriptomic data with known metabolic networks to investigate disease-associated metabolic changes. In this thesis, the mitochondrial compartment of HumanGEM was first curated through systematic comparison with the manually curated MitoCore model to im prove reaction directionality and compartment consistency. Longitudinal whole blood RNA-seq data from healthy control, prodromal PD, and established PD co horts were then integrated with the curated model using the ftINIT contextualization algorithm to generate 2326 patient-specific GEMs. Flux sampling was subsequently performed to characterize feasible metabolic states across patient-specific models. Global analyses of model structure and sampled flux-space organization revealed limited disease separation at the systems level. However, differential flux analysis identified metabolic alterations. Significant reactions were associated with mul tiple pathways that have been previously implicated in PD, such as amino acid metabolism, lysosomal degradation and fatty acid processing. Overall, this work demonstrates the application of patient-specific metabolic mod eling to investigate systemic metabolic alterations associated with Parkinson’s dis ease progression. Although the identified metabolic differences were modest and heterogeneous across longitudinal timepoints, the findings demonstrate the utility of GEM-based approaches for exploratory systems-level analysis of complex human disease