Single-cell biclustering for cell-specific transcriptomic perturbation detection in AD progression.

Area of Science:

• Genomics
• Computational Biology
• Neuroscience

Background:

• Alzheimer's disease (AD) pathogenesis involves complex, cell-specific gene regulatory alterations.
• Understanding these changes requires advanced computational tools for single-cell RNA sequencing (scRNA-seq) and single-nucleus RNA sequencing (snRNA-seq) data.

Purpose of the Study:

• To introduce single-cell Bayesian biclustering (scBC), a novel framework for identifying cell-specific gene network biomarkers.
• To analyze perturbations in functional gene modules at the single-cell level within AD datasets.
• To address common challenges in single-cell data analysis, such as batch effects and dropout events.

Main Methods:

• Developed the scBC framework utilizing Bayesian biclustering for gene network analysis.
• Applied scBC to Alzheimer's disease snRNA-seq data to identify perturbed gene modules.
• Incorporated prior biological knowledge to enhance interpretability.
• Evaluated scBC performance against state-of-the-art biclustering methods using simulated and real-world datasets.

Main Results:

• scBC successfully identified cell-specific gene network biomarkers in AD data.
• The framework revealed perturbations within gene modules across distinct cell populations during AD progression.
• scBC demonstrated robustness in handling batch effects and dropout events inherent in single-cell data.
• Comparative analyses confirmed the precision and reliability of scBC over existing methods.

Conclusions:

• scBC is a powerful tool for dissecting complex gene regulatory mechanisms in polygenic diseases like Alzheimer's.
• The framework enhances the biological interpretability of single-cell omics data.
• scBC offers a robust approach for unraveling intricate gene coexpression patterns in disease pathogenesis.