AtacAnnoR: a reference-based annotation tool for single cell ATAC-seq data.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Single-cell assay for transposase-accessible chromatin using sequencing (scATAC-seq) is crucial for understanding cell-type-specific chromatin accessibility.
• Accurate cell annotation is essential for interpreting scATAC-seq data, but remains challenging.
• Existing annotation methods often require paired scRNA-seq and scATAC-seq data.

Purpose of the Study:

• To develop and evaluate a novel computational method, AtacAnnoR, for annotating scATAC-seq data using reference scRNA-seq data.
• To assess the performance of AtacAnnoR against existing annotation tools on diverse benchmark datasets.
• To enhance annotation accuracy by incorporating strategies for utilizing multiple reference datasets.

Main Methods:

• AtacAnnoR employs a two-round annotation strategy.
• The method utilizes well-annotated scRNA-seq data as a reference for scATAC-seq data.
• A 'Combine and Discard' strategy is implemented to integrate annotations from multiple references.

Main Results:

• AtacAnnoR achieved the highest mean accuracy and balanced accuracy across 11 benchmark datasets compared to six competing methods.
• The method demonstrated robust performance when using unpaired scRNA-seq data as a reference.
• The 'Combine and Discard' strategy further improved annotation accuracy when multiple references were available.

Conclusions:

• AtacAnnoR provides a highly accurate and versatile method for annotating scATAC-seq data.
• The tool is particularly effective when only unpaired scRNA-seq data is available, broadening its applicability.
• Implemented as an R package, AtacAnnoR seamlessly integrates into existing scATAC-seq analysis pipelines.