Powerful and accurate detection of temporal gene expression patterns from multi-sample multi-stage single-cell...

Area of Science:

• Genomics
• Bioinformatics
• Statistical Modeling

Background:

• Single-cell RNA sequencing (scRNA-seq) enables high-resolution analysis of gene expression.
• Understanding dynamic gene expression changes over time is crucial in biological processes.
• Existing methods may not fully capture temporal dependencies in scRNA-seq data.

Purpose of the Study:

• To introduce TDEseq, a novel non-parametric statistical method for scRNA-seq time-course analysis.
• To leverage smoothing splines and mixed-effects models for accurate temporal pattern detection.
• To enhance the identification of dynamic gene expression changes within specific cell types.

Main Methods:

• Utilized smoothing splines basis functions to model time-dependent gene expression.
• Employed hierarchical structure linear additive mixed models to account for cellular correlations within individuals.
• Applied TDEseq to simulated datasets and four real scRNA-seq time-course studies.

Main Results:

• TDEseq successfully identified four distinct temporal gene expression patterns.
• The method demonstrated powerful performance in detecting dynamic expression changes.
• TDEseq achieved up to a 20% power gain compared to existing methods.
• Generated well-calibrated p-values, ensuring statistical reliability.

Conclusions:

• TDEseq offers a robust and powerful approach for analyzing temporal gene expression in scRNA-seq data.
• The method improves the detection of dynamic biological processes.
• TDEseq provides a valuable tool for researchers studying cellular dynamics over time.