A Pseudotime-Dependent TWAS Framework Identifies Disease Genes along Cell Developmental Paths.

Area of Science:

• Genetics and Genomics
• Computational Biology
• Systems Biology

Background:

• Transcriptome-wide association studies (TWAS) integrate gene expression and GWAS data for disease risk interpretation.
• Bulk-tissue expression data offers tissue-level insights, but single-cell data allows for finer cellular granularity.
• Existing single-cell methods may miss continuous cellular processes and misidentify causal cell stages.

Purpose of the Study:

• To develop a novel TWAS framework, pseudotime-dependent TWAS (pt-TWAS), for dissecting gene-disease associations at a finer cell-stage resolution.
• To capture continuous dynamic changes in gene expression along cell developmental paths.
• To improve statistical power and identify causal cell stages by modeling gene expression as a continuous function of pseudotime.

Main Methods:

• Developed the pt-TWAS framework modeling gene expression as a continuous function of pseudotime.
• Leveraged expression quantitative trait loci (eQTL) information across cell stages to boost statistical power.
• Constructed and visualized simultaneous confidence bands for gene effect curves to identify causal cell stages.

Main Results:

• pt-TWAS successfully replicated known risk genes in B-cell acute lymphoblastic leukemia (ALL).
• The method pinpointed specific cell stages associated with gene effects in ALL.
• pt-TWAS demonstrated statistical advantages over methods analyzing discrete cell types or stages.

Conclusions:

• pt-TWAS offers a powerful framework for fine-resolution analysis of gene-disease associations using single-cell data.
• The method accurately identifies causal cell stages by capturing continuous cellular processes.
• pt-TWAS enhances the understanding of genetic contributions to diseases at a cellular level.