scFPC-DE: Robust Differential Expression Analysis Along Single Cell Trajectories via Functional Principal Component...

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for understanding cellular development dynamics.
• Identifying temporally differentially expressed genes (TDEGs) aids in characterizing dynamic cellular states.
• Existing methods struggle with zero-inflation and ignore gene co-variation, leading to inaccuracies.

Purpose of the Study:

• To introduce scFPC-DE, a novel trajectory-based differential expression analysis method.
• To address limitations of existing methods, particularly false positives from zero-inflation in scRNA-seq data.
• To improve the power and interpretability of TDEG identification.

Main Methods:

• scFPC-DE utilizes functional data analysis (FDA) to model gene expression as a function of pseudotime.
• It employs functional principal component (FPC) analysis to capture gene expression covariance structure.
• The method mitigates the impact of zero inflation inherent in scRNA-seq data.

Main Results:

• scFPC-DE demonstrated superior control of type I errors and higher ROC-AUC in simulations.
• The method effectively captures informative gene expression patterns and shared variation along trajectories.
• Application to B cell data identified TDEGs relevant to differentiation pathways, outperforming existing methods.

Conclusions:

• scFPC-DE offers a robust approach for identifying TDEGs from scRNA-seq data.
• The method accurately captures temporal gene expression dynamics and reduces false positives.
• scFPC-DE enhances biological interpretability in cellular development studies.