Two-phase differential expression analysis for single cell RNA-seq

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) offers high-resolution transcriptome analysis.
• Traditional differential expression analysis often simplifies gene regulation to 'up or down' changes.
• scRNA-seq data allows for the observation of both binary (Off to On) and continuous (expression level) gene regulation.

Purpose of the Study:

• To develop a computational method that fully leverages single-cell data features for differential expression analysis.
• To identify and differentiate between binary and continuous gene regulation.
• To provide a more nuanced understanding of transcriptional regulation at the single-cell level.

Main Methods:

• Developed SC2P (Single-Cell 2-Phase) method for analyzing scRNA-seq data.
• SC2P identifies gene expression phases using model-based, data-driven approaches considering cell- and gene-specific contexts.
• The method distinguishes between 'phase transition' (binary regulation) and 'magnitude tuning' (continuous regulation).

Main Results:

• SC2P demonstrates substantial improvements in sensitivity for detecting differential expression compared to existing methods.
• The method maintains robust control over false discovery rates.
• SC2P offers enhanced interpretability of gene regulation types in single-cell transcriptomic data.

Conclusions:

• SC2P provides a more sensitive and robust approach to differential expression analysis in scRNA-seq data.
• The method enhances the understanding of gene regulation by distinguishing between phase transitions and magnitude tuning.
• SC2P advances the analysis of complex biological systems using single-cell transcriptomics.