Epimutation detection in the clinical context: guidelines and a use case from a new Bioconductor package

Area of Science:

• Genetics and Genomics
• Bioinformatics
• Computational Biology

Background:

• Epimutations, alterations in DNA methylation, are linked to rare diseases but pose detection challenges in clinical settings.
• Existing methods lack integration into standard pipelines or validation for rare disease data.
• Genome-wide epimutation detection using methylation microarrays is technically limited for clinical use.

Purpose of the Study:

• To develop and validate a Bioconductor package, epimutacions, for robust epimutation detection in rare diseases.
• To compare the performance of epimutacions against existing R packages like ramr.
• To provide guidelines for experimental design and data preprocessing for epimutation studies.

Main Methods:

• Developed the Bioconductor package 'epimutacions' implementing six statistical methods for epimutation detection.
• Created a user-friendly Shiny app for accessible epimutation analysis.
• Validated package performance using public datasets and analyzed large population cohorts (INMA, HELIX).

Main Results:

• Epimutacions demonstrated high performance, particularly with small sample sizes, outperforming ramr.
• Identified technical and biological factors influencing epimutation detection, offering practical guidelines.
• Most epimutations in general population cohorts did not correlate with significant gene expression changes.
• Successfully applied epimutacions to identify novel recurrent epimutations in autism candidate genes.

Conclusions:

• Epimutacions is a valuable tool for integrating epimutation detection into rare disease diagnostics.
• The package and associated app lower barriers for non-bioinformaticians in epimutation analysis.
• Provides essential guidelines for optimizing epimutation detection experiments and data analysis.