Abstract

Parkinson’s disease (PD) is a progressive neurodegenerative disorder marked by a prolonged prodromal phase and diverse motor and non-motor symptoms. Although clinical diagnostic criteria have improved, early diagnosis and monitoring of disease progression remain difficult due to the lack of accessible and reliable biomarkers. Metabolites and proteins from cerebrospinal fluid (CSF) and plasma have shown potential for distinguishing PD, but their utility in clinically deployable predictive models remains insufficiently characterized. We integrated proteomic and metabolomic profiles from CSF and plasma samples of over 1,100 participants in the Parkinson’s Progression Markers Initiative (PPMI) to identify biomarkers for early diagnosis and disease monitoring. Using univariate testing, interaction modeling, and multi-omics machine learning approaches, we characterized biofluid- and disease-specific signatures and evaluated their predictive performance. Multi-omics integration modeling identified 21 biomarker candidates validated across three models, with SVM and GLMNET achieving the highest recall of 86% and 83% and AUC values of 0.89–0.84 respectively. Longitudinal mixed effects modeling further revealed eight candidate biomarkers associated with disease progression. Including diagnostic markers: - Cadaverine, Secretogranin, Neuroendocrine protein 7B2, Limbic system-associated membrane protein. Prodromal to PD conversion: Neurosecretory protein VGF, Lumican. Sensitive progression markers: Kininogen-1, Galectin 3 binding protein. Collectively, these candidate biomarkers have direct clinical relevance, as they reflect complementary aspects of neurodegeneration, systemic inflammatory responses, and neurosecretory dysfunction implicated in PD. VGF and lumican show promise for identifying patients at the critical prodromal-to-PD conversion phase; diagnostic markers like cadaverine, neuroendocrine protein 7B2, secretogranin, and limbic system-associated membrane protein may be used to support early disease detection.  Progression-sensitive markers may allow for longitudinal tracking of disease trajectories. This can be translated into a foundation for stratified patient management and biomarker-informed therapeutic treatments.

School

School of Sciences and Engineering

Department

Institute of Global Health & Human Ecology

Degree Name

PhD in Applied Sciences

Graduation Date

Fall 2-15-2026

Submission Date

1-25-2026

First Advisor

Ahmed Moustafa

Second Advisor

Mohamed Salama

Committee Member 1

Ramy Aziz

Committee Member 2

Mohamed El-Hadidi

Extent

178p.

Document Type

Doctoral Dissertation

Institutional Review Board (IRB) Approval

Approval has been obtained for this item

Disclosure of AI Use

Thesis editing and/or reviewing

Download

Available for download on Tuesday, January 25, 2028

Share

COinS