Profiling the Braak progression in Parkinson's disease: a transcriptomics and ML driven identification of progressive...

Parkinson's disease (PD), the second most prevalent neurodegenerative disorder, is marked by dopaminergic neuron loss and α-synuclein aggregation. Although Braak staging based on Lewy body pathology stratifies disease progression, clinical variability persists. Identifying stage-specific molecular signatures is essential for developing effective progressive biomarkers. Here, we integrated midbrain transcriptomic data from microarray, bulk RNA-seq, and snRNA-seq platforms. Subsequently, to catalogue candidate biomarkers of PD progression, we employed a three-layered approach integrating Protein-protein-interaction networks, shortest-path analysis, and ModuLand-based module detection. By mapping gene interactions within and across Braak stages, we identified core module genes that influence network topology and play pivotal roles in disease advancement. Braak stage transition analysis further highlighted 66 key genes, significantly enriched in ubiquitin-mediated proteolysis, cytokine signaling, and neurodegenerative pathways. ElasticNet regression short-listed these to ten candidate biomarkers with strong predictive power (AUC ≥ 0.90), and six genes consistently showed good accuracy (AUROC > 0.75) across Random Forest, GBM, and CatBoost models. Validation with blood transcriptome data supported their utility as non-invasive biomarker. Further, survival analysis of PPMI data highlighted the prognostic significance of HSPA1B, LRRK2, and DNAJB1, identifying them as potential risk associated genes in PD. Among these, only HSPA1B yielded a competitive-endogenous-RNA network with prognostic relevance, and its differential expression between early and late PD stages further supported its role as a progressive biomarker. The key regulatory axis identified-XIST → miR-221-3p/miR-23a-5p/miR-548au-3p → HSPA1B offers both prognostic value and mechanistic insight into PD progression, with potential for early diagnosis and targeted therapy.