A Hybrid Machine Learning and Network Analysis Approach Reveals Two Parkinson's Disease Subtypes from 115 RNA-Seq Post-Mortem Brain Samples

Abstract

Precision medicine emphasizes fine-grained diagnostics, taking individual variability into account to enhance treatment effectiveness. Parkinson’s disease (PD) heterogeneity among individuals proves the existence of disease subtypes, so subgrouping patients is vital for better understanding disease mechanisms and designing precise treatment. The purpose of this study was to identify PD subtypes using RNA-Seq data in a combined pipeline including unsupervised machine learning, bioinformatics, and network analysis. Two hundred and ten post mortem brain RNA-Seq samples from PD (n = 115) and normal controls (NCs, n = 95) were obtained with systematic data retrieval following PRISMA statements and a fully data-driven clustering pipeline was performed to identify PD subtypes. Bioinformatics and network analyses were performed to characterize the disease mechanisms of the identified PD subtypes and to identify target genes for drug repurposing. Two PD clusters were identified and 42 DEGs were found (p adjusted ≤ 0.01). PD clusters had significantly different gene network structures (p < 0.0001) and phenotype-specific disease mechanisms, highlighting the differential involvement of the Wnt/β-catenin pathway regulating adult neurogenesis. NEUROD1 was identified as a key regulator of gene networks and ISX9 and PD98059 were identified as NEUROD1-interacting compounds with disease-modifying potential, reducing the effects of dopaminergic neurodegeneration. This hybrid data analysis approach could enable precision medicine applications by providing insights for the identification and characterization of pathological subtypes. This workflow has proven useful on PD brain RNA-Seq, but its application to other neurodegenerative diseases is encouraged.

Keywords: Parkinson’s disease; RNA-seq; data science; genomic data science; machine learning; network analysis; precision medicine; subtyping.

Figure 1

Clustered subjects and clinical characterization. (A) PCA plot for RNA-Seq data, clearly showing PDC1 and PDC2 separation. (B) Boxplot for PDC1 and PDC2 age at death, reporting a significant difference * p value < 0.05.

Figure 2

Differences between PDC1 and PDC2 gene expression. (A) Volcano plot reporting differences in gene expression where red points were DEGs. (B) Gene network for PDC1 showing nodes are segregated into 2 communities. (C) Gene network for PDC2, showing no segregation in communities. Both networks were built with the Kamada–Kawai layout, where each node is a gene and each edge is a PCIT value between genes.

Figure 3

Differential network and measures of network regulators. (A) Differential network for PDC1 and PDC2 DEGs networks. (B) Bar plot for standardized centrality of genes from the differential network. (C) RIF 1 scores of genes from the differential network. (D) Difference in standardized centrality between genes from PDC1 and PDC2 networks (shown in Figure 2B,C).

Figure 4

Venn diagram reporting the number of unique and overlapping DEGs found in PDC1 vs. NC and PDC2 vs. NC comparisons.

Figure 5

Investigation of the disease mechanism of PDC1, based on PDC1 vs. NC unique DEGs. (A) Enrichment analysis of DEGs shows that these genes map mostly on synaptic functions, in particular on the glutamatergic synapse. (B) Pheatmap illustration of the glutamatergic synapse pathway, showing downregulation in neurotransmitter uptake (vGLUT) and feedback regulation of neurotransmission (mGluR2) functions.

Figure 6

Investigation of the disease mechanism of PDC2, based on PDC2 vs. NC unique DEGs. (A) Enrichment analysis of DEGs shows that these genes map on inflammatory processes. (B) Gene network showing segregation in communities (only communities with significant PPI scores are colored). The network was built with the Kamada–Kawai layout, where each node is a gene and each edge is a PCIT value between genes.

Figure 7

PRISMA flow diagram showing the step-by-step process of our search and selection applied during systematic data retrieval.