Best practices for differential accessibility analysis in single-cell epigenomics.

Area of Science:

• Single-cell epigenomics
• Computational biology
• Genomics

Background:

• Differential accessibility (DA) analysis is crucial for understanding cell identity and responses.
• Existing statistical methods for DA analysis in single-cell epigenomics lack clear performance guidelines.
• There is no consensus on the optimal statistical approaches for scATAC-seq data.

Purpose of the Study:

• To systematically evaluate the performance of statistical methods for identifying DA regions in single-cell ATAC-seq (scATAC-seq) data.
• To assess the accuracy, bias, robustness, and scalability of various DA analysis methods.
• To establish best practices for scATAC-seq data analysis and develop a corresponding R package.

Main Methods:

• Leveraged a compendium of scATAC-seq experiments.
• Utilized matching bulk ATAC-seq or scRNA-seq data for validation.
• Systematic evaluation of statistical methods based on accuracy, bias, robustness, and scalability.

Main Results:

• Identified key principles governing the performance of DA analysis methods.
• Provided a comprehensive assessment of commonly used statistical approaches for scATAC-seq.
• Developed an R package to implement recommended best practices for DA analysis.

Conclusions:

• The study clarifies the performance characteristics of different DA analysis methods for scATAC-seq data.
• Best practices for scATAC-seq DA analysis have been defined and implemented in a user-friendly R package.
• This work facilitates more reliable and reproducible discovery of regulatory programs from single-cell epigenomic data.