Analyzing spatial transcriptomics and neuroimaging data in neurodegenerative diseases

Neurodegenerative diseases are hallmarked by protein inclusions and cell loss in disease-related brain regions, but the molecular mechanisms that lead to the pathological and symptomatic hallmarks of neurodegeneration are still not fully understood. In this thesis, we make use of bioinformatics approaches to analyze a high-resolution spatial gene expression atlas of the healthy human brain generated by the Allen Institute of Brain Science. Spatial transcriptomics allows examining the molecular and functional organization of the human brain and can be combined with neuroimaging data to identify brain regions and anatomical structures that are vulnerable to cell loss in neurodegenerative diseases. By combining both data modalities, we examined healthy molecular functions in brain regions associated with disease vulnerability based on neuroimaging features, namely gray matter loss within brain networks in individuals with Parkinson’s disease, Huntington’s disease, and individuals at risk of schizophrenia. With this thesis, we have shown that by applying data-driven computational methods we can explore the whole genome and find gene expression patterns informative of regional brain vulnerability in neurodegenerative diseases. Our methods can similarly be applied to unravel the molecular mechanisms in other neurodegenerative diseases, and potentially even reveal shared mechanisms between neurological disorders.