Overcoming confounding plate effects in differential expression analyses of single-cell RNA-seq data

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) is a powerful tool for analyzing gene expression at the individual cell level.
• Detecting differentially expressed (DE) genes is a common analysis for scRNA-seq data.
• Plate effects, arising from separate experimental processing, often confound biological groups in scRNA-seq experiments and are frequently overlooked.

Purpose of the Study:

• To investigate the impact of unaddressed plate effects on DE gene analysis in scRNA-seq data.
• To propose and validate a method to mitigate confounding plate effects in scRNA-seq DE analyses.

Main Methods:

• Simulated scRNA-seq data with known plate effects were used to evaluate DE analysis methods.
• A novel approach involving summing gene counts across cells within each plate prior to DE analysis was proposed.
• The proposed summation method was compared against standard DE analysis ignoring plate effects using simulated and real scRNA-seq datasets.

Main Results:

• Failure to account for plate effects in DE analyses leads to a loss of type I error control.
• Summing gene counts per plate before DE analysis effectively restores type I error control in the presence of plate effects.
• The summation method demonstrated robustness to variations in cell numbers and library sizes across plates.
• Analysis of real scRNA-seq data using the summation method showed improved specificity and gene ranking.

Conclusions:

• Ignoring plate effects in scRNA-seq DE analysis compromises statistical validity.
• Summing gene counts per plate is a robust and effective strategy to maintain statistical rigor in scRNA-seq DE analyses with confounding plate effects.
• This approach enhances the reliability of identifying differentially expressed genes in experiments with batch variations.