PETScan: score-based genome-wide association analysis of RNA-Seq and ATAC-Seq data

Area of Science:

• Systems biology
• Genomics
• Bioinformatics

Background:

• High-dimensional sequencing data like RNA-Seq (gene expression) and ATAC-Seq (chromatin accessibility) are crucial for systems biology.
• Current methods often overlook long-range cis-regulatory interactions and fail to adequately model count-based sequencing data.
• Understanding gene regulation requires analyzing associations between gene expression and chromatin accessibility.

Purpose of the Study:

• To develop a computationally efficient method for genome-wide association analysis of RNA-Seq and ATAC-Seq data.
• To address limitations of existing tools by incorporating distant regulatory element interactions and accommodating count-based data.
• To provide a robust tool for uncovering gene regulatory mechanisms.

Main Methods:

• Introduced PETScan, a PEak-Transcript Score-based association analysis method.
• Utilized negative binomial models to accurately represent RNA-Seq count data.
• Employed score tests and matrix calculations for computational efficiency, combined with empirical permutation and genomic control for valid p-value calculations.

Main Results:

• PETScan demonstrated computational efficiency, achieving speeds three orders of magnitude faster than Wald tests.
• The method identified significant gene-peak pairs comparable to existing approaches.
• PETScan effectively analyzes genome-wide associations between gene expression and chromatin accessibility.

Conclusions:

• PETScan offers a computationally efficient and statistically robust approach for analyzing high-dimensional sequencing data.
• The tool enhances the study of gene regulatory mechanisms by considering both local and distant interactions.
• PETScan provides valuable insights for systems biology research and is available as an R package.