A statistical framework for differential pseudotime analysis with multiple single-cell RNA-seq samples.

Area of Science:

• Computational Biology
• Genomics
• Single-cell Analysis

Background:

• Pseudotime analysis of single-cell RNA sequencing (scRNA-seq) data is crucial for understanding dynamic gene regulatory programs.
• Existing methods struggle to compare pseudotemporal patterns across multiple samples and experimental conditions.
• Comparing pseudotemporal patterns across samples is essential for robust biological discovery.

Purpose of the Study:

• To introduce Lamian, a novel computational framework for differential multi-sample pseudotime analysis.
• To enable statistically rigorous comparison of biological processes across different samples and conditions.
• To identify changes in gene expression, cell density, and trajectory topology in a multi-sample context.

Main Methods:

• Development of Lamian, a comprehensive computational framework for differential multi-sample pseudotime analysis.
• Statistical inference accounting for cross-sample variability to reduce false discoveries.
• Application to real scRNA-seq data and simulation data, including COVID-19 patient immune responses.

Main Results:

• Lamian identifies changes in biological processes associated with sample covariates (e.g., disease severity).
• The framework effectively adjusts for batch effects in multi-sample scRNA-seq data.
• Lamian demonstrates reduced sample-specific false discoveries compared to existing methods.

Conclusions:

• Lamian provides a statistically rigorous approach for differential multi-sample pseudotime analysis.
• The framework enhances the ability to decode cellular gene expression programs in continuous biological processes.
• Lamian is advantageous for analyzing dynamic biological processes across multiple samples and conditions.