scAnnoX: an R package integrating multiple public tools for single-cell annotation.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Single-cell annotation is vital for single-cell genomics data analysis.
• A lack of integrated tools hinders the comparison of existing single-cell annotation algorithms.

Purpose of the Study:

• To develop a comprehensive R package for integrating and comparing single-cell annotation algorithms.
• To evaluate the performance of ten widely adopted single-cell annotation algorithms.

Main Methods:

• Ten single-cell annotation algorithms were selected based on reference dataset-dependent or marker gene-dependent approaches.
• Algorithms evaluated include SingleR, Seurat, sciBet, scmap, CHETAH, scSorter, sc.type, cellID, scCATCH, and SCINA.
• An R package, scAnnoX, was developed for integration and comparative analysis.

Main Results:

• scAnnoX provides a unified framework for cell annotation in scRNA-seq data.
• It streamlines the testing, evaluation, and comparison of annotation algorithms.
• SingleR, Seurat, sciBet, and scSorter showed high prediction accuracy, with SingleR and sciBet being top performers.

Conclusions:

• scAnnoX facilitates efficient comparative analysis of cell type annotations in scRNA-seq datasets.
• The package guides users towards high-performing algorithms like SingleR and sciBet.
• The scAnnoX package is available at https://github.com/XQ-hub/scAnnoX.