Supervised latent factor modeling isolates cell-type-specific transcriptomic modules that underlie Alzheimer's...

Area of Science:

• Neuroscience
• Genomics
• Computational Biology

Background:

• Late-onset Alzheimer's disease (AD) involves progressive neurodegeneration and brain atrophy, with early changes preceding symptoms.
• While neuronal loss is a hallmark, recent studies emphasize the critical role of glial cells, particularly microglia, in AD pathogenesis.
• Genome-wide association studies (GWAS) and single-nucleus RNA sequencing (snRNA-seq) have implicated glial cells in AD.

Purpose of the Study:

• To apply pattern-learning algorithms to integrated transcriptomic data to identify Alzheimer's disease (AD)-predictive gene modules across major brain cell types.
• To determine the biological relevance of these modules by identifying enriched signaling pathways.
• To infer disease progression trajectories and quantify cell-type module interactions in the AD brain.

Main Methods:

• Utilized pattern-learning algorithms on whole-transcriptome data from brain samples.
• Integrated snRNA-seq data to identify distributed gene modules.
• Applied module predictions to infer disease pseudo-trajectory and validated with post-mortem tissue markers.
• Quantified interactions between cell-type specific modules and localized AD risk genes.

Main Results:

• Identified biologically meaningful, AD-predictive gene modules within various brain cell types, notably microglia.
• Demonstrated the predictive power of these modules to infer disease progression along a pseudo-trajectory.
• Confirmed module relevance through enriched signaling cascades and identified novel AD-related pathways.
• Quantified inter-module crosstalk and mapped known AD risk genes to specific module gene programs.

Conclusions:

• Advocates for a shift from cell-type-specific analysis to gene module specificity for a deeper understanding of AD.
• Highlights the potential of specific gene programs, particularly in microglia, for predicting disease trajectory.
• Suggests that focusing on gene modules can refine the understanding of genome-wide AD risk loci and their functional roles.