Identification of differentially methylated loci using wavelet-based functional mixed models.

Area of Science:

• Epigenetics and Genomics
• Computational Biology
• Statistical Genetics

Background:

• DNA methylation is a crucial epigenetic modification regulating gene expression, with alterations linked to complex diseases like cancer.
• Previous studies often focused on limited regions, but genome-wide analysis is now essential for comprehensive understanding.
• Identifying differentially methylated loci is key to understanding disease mechanisms.

Purpose of the Study:

• To introduce and evaluate a novel wavelet-based functional mixed model (WFMM) for analyzing high-throughput genome-wide DNA methylation data.
• To improve the detection power of differentially methylated loci compared to methods that ignore spatial correlations.
• To provide a flexible framework applicable to diverse methylation profiling datasets.

Main Methods:

• Application of a wavelet-based functional mixed model (WFMM) methodology.
• Analysis of high-throughput methylation data, accommodating spatial correlations across the genome and correlations between samples.
• Basis function modeling and functional random effects to capture complex data structures.

Main Results:

• A simulation study demonstrated that the WFMM approach outperforms independent locus modeling in detecting differentially methylated loci.
• Application to the THREE dataset identified novel regions of differential methylation not previously reported.
• The WFMM framework successfully analyzed diverse datasets including CpG Shore, THREE, and NIH Roadmap Epigenomics data.

Conclusions:

• The WFMM methodology offers a powerful and flexible approach for genome-wide differential DNA methylation analysis.
• This method enhances the detection of biologically relevant methylation changes, potentially leading to new insights in complex diseases.
• The WFMM software and associated datasets are publicly available for broader research application.