Disentangling associations between complex traits and cell types with seismic.

Area of Science:

• Genomics
• Computational Biology
• Neuroscience

Background:

• Integrating single-cell RNA sequencing (scRNA-seq) with Genome-Wide Association Studies (GWAS) is crucial for understanding complex traits and diseases.
• Existing methods often struggle with scalability, interpretability, and robustness in analyzing scRNA-seq and GWAS data.
• Identifying specific cell types and genes involved in polygenic traits remains a significant challenge.

Purpose of the Study:

• To introduce seismic, a novel computational framework for integrating scRNA-seq and GWAS data.
• To develop a new specificity score and influential gene analysis for robust cell type-trait association.
• To enhance the discovery of cell-type-specific biological mechanisms underlying complex traits and diseases.

Main Methods:

• Developed seismic, a framework incorporating a novel specificity score (expression magnitude and consistency) and influential gene analysis.
• Applied seismic to over 1000 cell-type characterizations across various granularities and 28 polygenic traits.
• Utilized pathology-based Alzheimer's GWAS data to identify vulnerable neuron populations and molecular pathways.

Main Results:

• Seismic successfully corroborates known trait-cell type associations and identifies novel relevant cell groups.
• The framework reveals cell- and brain-region-specific pathological differences in Parkinson's and Alzheimer's disease.
• Identified vulnerable neuron populations and implicated molecular pathways in Alzheimer's neurodegeneration using pathology-based GWAS.

Conclusions:

• Seismic provides a computationally efficient, powerful, and interpretable method for mapping polygenic trait relationships to cell-type-specific expression.
• The framework offers new insights into disease mechanisms by dissecting cell-type and regional specificity.
• Seismic advances the integration of multi-omics data for biological discovery and precision medicine.