Transcriptomics profiling of Parkinson’s disease progression subtypes reveals distinctive patterns of gene expression

The present PhD dissertation reports a research project on Parkinson’s Disease (PD) subtyping. Parkinson’s Disease (PD) exhibits considerable heterogeneity, necessitating the identification of subtypes for personalized therapeutic strategies. In this series of studies, a few experiments have been conducted to answer questions about PD subtypes using transcriptomics data. RNA-Seq data from Parkinson’s Progression Markers Initiative (PPMI) was used, allowing the application of advanced data analytics methods on one of the largest PD RNA samples available to date, with more than 407 PD subjects and 193 Healthy Control (HC) subjects. Results from the first two experiments aiming to detect the existence of clusters of subjects considering transcriptomics data did not evidence any disease subgroup, neither at the time of diagnosis through cross-sectional data, nor by change over time through longitudinal data. These findings brought to design the main experiment reported in the present thesis. Artificial Intelligence (AI) has recently helped to identify three PD progression subtypes using clinical data, but their transcriptomics profiles remained unexplored. This study aimed at identifying the transcriptomics characteristics of PD progression subtypes and assessing the utility of gene expression data in subtype prediction at baseline. In particular, its objectives were (1) to describe the transcriptomics characteristics of three PD progression subtypes identified using AI, and (2) to subsequently evaluate the usefulness of gene expression data in predicting disease subtype at baseline. Whole blood RNA-Sequencing data underwent differential gene expression analysis, followed by extensive analysis of the associated biological pathways. A Machine Learning (ML) classifier was trained using data from multiple modalities, including gene expression values. Results included Differentially Expressed Genes (DEGs) uniquely associated with the progression subtypes, which were, notably, distinct TRANSCRIPTOMICS OF PD SUBTYPES Contents from commonly found DEGs in PD transcriptomics studies. Pathway analyses showed both distinct and shared characteristics among subtypes, with two of them having opposite expression patterns for pathways involved in immune response alterations, and the third having a more unique profile with respect to the others, characterized by increased expression of genes related to detoxification processes. All three subtypes showed a significant modulation of pathways related to the regulation of gene expression, metabolism, and cell signaling. ML revealed that the progression subtype with the worst prognosis can be predicted at baseline with 0.877 AUROC, yet the contribution of gene expression was marginal for the prediction of the subtypes. The present study provides novel information regarding the transcriptomics profiles of PD progression subtypes, which may foster precision medicine providing relevant indications for a finer-grained diagnosis and prognosis.