The Quest for a Universal Parkinson’s Transcriptomics Signature is Derailed by Inherent Variability between Patients

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Objective

To systematically evaluate the reproducibility and clinical utility of published blood mRNA-based gene signatures for classifying PD from healthy controls, and to uncover the pitfalls that limit their performances.

Methods

We validated the classification performance of 13 gene signatures unique for PD (published 2015-2025) using the Parkinson’s Progression Markers Initiative (PPMI) database. We further validated classification performance on data collected in a prospective clinical trial where demographic and clinical parameters were minimized and environmental confounders were strictly controlled. Sources of gene expression variability were studied by mean pairwise distance analysis.

Results

Gene overlap between signatures was low (11/411, 2.7%) but statistically significant ( P <0.001) and enriched with genes involved in lipid metabolism. The majority (10/13) of these signatures retained statistical significance when tested on the PPMI dataset, but their classification performance was modest (median AUC 59.7%), substantially lower than originally reported. Classification performance improved when comparing GBA1 -associated PD to controls (median AUC 65.4%, P =0.006). A prospective trial (16 PD patients, 14 controls) with rigorous environmental standardization did not improve classification accuracy. Inter-individual expression variability is the dominant limiting factor in classifier performance. Variability was not impacted by long-term levodopa therapy.

Interpretation

While blood transcriptomic signatures unique to PD are reproducible and may elucidate PD pathophysiology, their broad clinical utility remains severely limited by inherent inter-individual variability. Future research should prioritize biomarkers of divergence rather than the pursuit of a universal PD signature.

Article activity feed