Differential methylation analysis for BS-seq data under general experimental design.

Area of Science:

• Epigenetics and Genomics
• Computational Biology
• Statistical Genetics

Background:

• DNA methylation is a key epigenetic regulator in biological processes and diseases.
• Bisulfite sequencing (BS-seq) is a high-resolution technology for profiling DNA methylation.
• Existing statistical methods for differential methylation analysis are limited to two-group comparisons.

Purpose of the Study:

• To develop a flexible and powerful statistical method for detecting differential DNA methylation.
• To address the limitations of current methods in handling general and multifactor experimental designs in BS-seq data.
• To provide a computationally efficient tool for analyzing complex BS-seq experiments.

Main Methods:

• A novel beta-binomial regression model with an arcsine link function was developed.
• Parameter estimation utilizes a generalized least squares approach on transformed data.
• The method avoids iterative algorithms, enhancing computational efficiency.

Main Results:

• The proposed method accurately detects differentially methylated loci under general experimental designs.
• Simulations and real data analyses confirm the method's power and robustness.
• The approach is computationally efficient, suitable for large-scale BS-seq studies.

Conclusions:

• The novel statistical model offers an accurate, powerful, and efficient solution for differential methylation analysis in BS-seq data.
• This method supports complex experimental designs beyond simple two-group comparisons.
• The tool is available as the Bioconductor package DSS for broader accessibility.