Correlation Networks To Uncover Changes in Protein Relationships in Spinocerebellar Ataxia Type 2 and Cerebellar Multiple System Atrophy

Rare neurological diseases (RNDs) are complex diseases characterized by significant heterogeneity in genetic, molecular, and pathological characteristics, which make them poorly understood and still challenging to diagnose and treat. Therefore, identifying methods that help improve diagnosis and discrimination efficiency is crucial. In this context, patients presenting overlapping phenotypes and different etiologies may offer the opportunity to study disease-specific pathways and differential characteristics in fluid biomarkers. In this study, a label-free proteomic analysis was set up to capture the protein profile of peripheral blood mononuclear cells (PBMCs) isolated from the plasma of adult patients diagnosed with multiple system atrophy-cerebellar subtype (MSA-C) or spinocerebellar ataxia type 2 (SCA2), and matched healthy controls (CTR). Gaussian graphical models and graphical LASSO were implemented as a novel data analysis approach for discriminating between the two diseases. This approach allowed us to identify coregulation networks that are different between the diseases and the controls. Most importantly, this work introduces an innovative workflow for proteomics research that might be employed as a complement to traditional methods.