Bigmelon: tools for analysing large DNA methylation datasets.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• DNA methylation datasets are rapidly increasing in size.
• Large datasets from platforms like HumanMethylationEPIC exceed R's memory capacity.
• There is a critical need for computationally efficient methods for high-dimensional data analysis.

Purpose of the Study:

• Introduce the bigmelon R package for memory-efficient DNA methylation data analysis.
• Provide a practical workflow for large-scale epigenome-wide association studies (EWAS).
• Enable complex analyses without requiring substantial RAM.

Main Methods:

• Utilizes the CoreArray Genomic Data Structure (GDS) file format.
• Leverages the gdsfmt package for efficient data handling.
• Implements a workflow for reading, preprocessing, quality control, and statistical analysis of DNA methylation data.

Main Results:

• Demonstrates the bigmelon package's capability with a large dataset of 1193 human blood samples.
• Successfully handles complex, large-scale analyses required for EWAS.
• Facilitates analysis without prohibitive memory requirements.

Conclusions:

• The bigmelon R package provides a computationally efficient solution for analyzing large DNA methylation datasets.
• It enables researchers to conduct complex EWAS analyses on high-dimensional data.
• The package is available on Bioconductor, promoting accessibility.