Spatially Enhanced Differential RNA Methylation Analysis from Affinity-Based Sequencing Data with Hidden Markov...

Area of Science:

• Epigenetics and Genomics
• RNA Biology
• Bioinformatics

Background:

• Next-generation sequencing enables profiling of the N6-methyl-adenosine (m6A) RNA methylome using methylated RNA immune-precipitation sequencing (MeRIP-Seq).
• MeRIP-Seq, while powerful, lacks base-pair resolution, often identifying single sites that contain multiple, differentially methylated RNA residuals.
• Existing peak-based methods struggle to distinguish these multiple residuals within a single site.

Purpose of the Study:

• To develop a novel computational approach for high-resolution analysis of MeRIP-Seq data.
• To accurately differentiate multiple RNA methylation residuals within a single MeRIP-Seq site.
• To improve the detection of differential RNA methylation between biological conditions, such as normal versus cancerous tissues.

Main Methods:

• Proposed a hidden Markov model (HMM) based algorithm to analyze MeRIP-Seq data.
• Divided MeRIP-Seq detected methylation sites into smaller, adjacent bins.
• Applied the HMM to model dependencies between spatially adjacent bins for enhanced resolution and accuracy.

Main Results:

• The HMM-based algorithm demonstrated superior performance compared to existing peak-based methods on simulated data.
• The proposed method achieved higher statistical significance in detecting differential methylation regions on real biological datasets.
• Successfully differentiated multiple RNA methylation residuals within single sites, a limitation of previous approaches.

Conclusions:

• The developed HMM approach significantly enhances the resolution and accuracy of MeRIP-Seq data analysis.
• This method offers a more precise tool for investigating differential RNA methylation, crucial for understanding disease mechanisms.
• Provides a statistically robust framework for identifying subtle yet significant epigenetic changes in RNA.