Characterization of gene expression patterns in mild cognitive impairment using a transcriptomics approach and neuroimaging endophenotypes

INTRODUCTION: Identification of novel therapeutics and risk assessment in early stages of Alzheimer's disease (AD) is a crucial aspect of addressing this complex disease. We characterized gene‐expression patterns at the mild cognitive impairment (MCI) stage to identify critical mRNA measures and gene clusters associated with AD pathogenesis. METHODS: We used a transcriptomics approach, integrating magnetic resonance imaging (MRI) and peripheral blood‐based gene expression data using persistent homology (PH) followed by kernel‐based clustering. RESULTS: We identified three clusters of genes significantly associated with diagnosis of amnestic MCI. The biological processes associated with each cluster were mitochondrial function, NF‐kB signaling, and apoptosis. Cluster‐level associations with cortical thickness displayed canonical AD‐like patterns. Driver genes from clusters were also validated in an external dataset for prediction of amyloidosis and clinical diagnosis. DISCUSSION: We found a disease‐relevant transcriptomic signature sensitive to prodromal AD and identified a subset of potential therapeutic targets associated with AD pathogenesis.