scMerge leverages factor analysis, stable expression, and pseudoreplication to merge multiple single-cell RNA-seq...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) generates high-dimensional data for cellular analysis.
• Integrating multiple scRNA-seq datasets can reveal biological insights beyond individual studies.
• Challenges exist in merging datasets due to technical variations and batch effects.

Purpose of the Study:

• To introduce scMerge, a novel algorithm for integrating multiple scRNA-seq datasets.
• To demonstrate scMerge's effectiveness in improving cell type separation and biological discovery.
• To provide a robust method for enhancing the analysis of large-scale scRNA-seq data.

Main Methods:

• scMerge utilizes factor analysis of stably expressed genes.
• Pseudoreplicates are employed across datasets for robust integration.
• The algorithm was benchmarked against existing methods using public scRNA-seq datasets.

Main Results:

• scMerge consistently outperformed published methods in cell type separation.
• The algorithm effectively removes unwanted technical factors from integrated data.
• scMerge facilitated the inference of developmental trajectories in a liver dataset collection.

Conclusions:

• scMerge offers a powerful approach for robustly integrating multiple scRNA-seq datasets.
• The algorithm enhances biological discovery by improving data quality and interpretability.
• scMerge is a valuable tool for researchers analyzing large-scale single-cell data.