scMatch: a single-cell gene expression profile annotation tool using reference datasets.

Area of Science:

• Computational biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) provides gene expression data at the individual cell level.
• Accurate cell identity inference is crucial for analyzing complex tissues using scRNA-seq.
• Current clustering-based methods struggle with low-coverage scRNA-seq data.

Purpose of the Study:

• To introduce scMatch, a novel tool for direct single-cell annotation.
• To evaluate scMatch's performance against existing methods and assess influencing factors.
• To enable precise cell population identification in complex tissues using customizable reference datasets.

Main Methods:

• Developed scMatch for direct single-cell annotation by matching to reference datasets.
• Evaluated scMatch using various single-cell datasets, assessing sequencing depth, similarity metrics, and reference set composition.
• Tested scMatch's ability to integrate multiple data sources into customized reference profiles.

Main Results:

• scMatch rapidly and robustly annotates single cells.
• Achieved comparable accuracy to SingleR but with increased speed and scalability.
• Demonstrated effective annotation using large, customized reference datasets from multiple sources.

Conclusions:

• scMatch provides an efficient and accurate method for single-cell annotation.
• The tool's scalability and flexibility empower researchers to analyze diverse complex tissues.
• scMatch facilitates precise identification of cell populations in transcriptomic data.