CpG Island shelves

Nowadays, regions with relatively lower CpG density are gaining importance in DNA methylation studies. This is based on the fact that a majority of CpG rich regions (CpG islands) are non-dynamic and less variant in terms of methylation status probed across a variety of tissues and cell populations (Irizarry 2009, Ziller 2013). It is now proven that methylation is more dynamic along the CpG shores (< 2kb flanking CpG Islands) and CpG shelves (<2kb flanking outwards from a CpG shore). While it is easy to retreive the genomic co-ordinates of CpG Islands from UCSC browser, public resources for retrieving the genomic co-ordinates for shores and shelfs were missing or scarce. Here, I am displaying the code (in R using Bioconductor package GenomicRanges) I use for generating objects for CpG islands, CpG island shores and CpG island shelves. The resulting objects are GRanges objects which could be used in multiple downstream applications in Bioconductor related packages.

Scope of Reduced Representation Bisulfite Sequencing Data

RRBS method explores the methylation status across the genome but at specific locations defined by the MspI recognition sites. These sites are mostly located within CpG Islands. So, how to visulaize the scope of the RRBS data - regions from where we can expect the methylation status in the human genome.

I show the localization of the CpG Island on the human genome in the following graphic. These are the most possible locations for the RRBS data sampling

This is a Karyogram view of the CpG Island on human chromosomes. Chromosomes were plotted relative to their size. Each CpG island is denoted by a single dot at the relative position on the chromosome. CpG Island close to each other are perceived as a connecting line to the human eye (hundred of dot placed closely). This picture also gives us an Idea that certain parts of the chromosomes either lack CpG islands (chr1, chr13, chr14, chr15). This could be due to incomplete mapping of human genome on these chromosomes.

I made this graphic using ggbio and ggplot2.

Methylation Status of CpG Islands Across Human Genome

Researchers are aware that a majority of the CpG islands are unmethylated. How to represent this fact in a graphic? 

The above picture created with ggplot2 explains us how the CpG methylation across the CpG Islands is distributed for each chromosome. Each CpG island is shown as a single dot(.). Methylation on the CpG island is identified by the color gradient. This explains that most of the CpG Islands are unmethylated (overlapped dots are seen as a line on the left side of the picture). Sparsely methylated CpG Islands can be identified as blue dots on the right side. Click the picture for a larger view.