Journal of Computer Science & Systems Biology

Binary Transition Pattern (BTP) for Erythropoietic DNA Methylation Data

Abstract

Najyah Garoot* and Byung Guk Kim

Objective: Despite whole-genome data analysis for understanding the implications of differential DNA methylation patterns and the nature of their correlation with gene expression (GE) regulation, the observations remain controversial.

Methods: Unlike the usual methods of selecting differentially methylated regions (DMR) by hypothesis statistical testing, we clustered CpG sites by their actual log2FC over time. Each cluster is represented by a 5-digit binary code.

Results: We identified novel stage-specific TFBS during erythropoiesis. Our stage-specific differential methylation clusters have the ability to determine TFBS of co-regulators such as GATA1 and PU.1 at hypermethylated CpG loci.

Conclusion: Hypomethylated CpG loci clusters that are associated with GE upregulation are independent of hypermethylated sites associated with downregulation. The analy-sis confirms that DNA methylation has unidirectional correlation with GE.