Hi guys:
I'm working with whole-genome bisulfite sequencing DNA methylation data (mainly interested in CpG methylation). I have a low number of samples, and while I have found extensive methods to define differentially methylated regions (DMR) by comparing groups of samples, or even 1 vs 1 settings, I was wondering if there are software packages to define methylation regions in a single sample.
(This is the most comprehensive review I have found on DMR-finding packages).
I want to independently describe the methylome of individual samples, so as to define methylation blocks, or regions of nearby cytosines displaying similar and consistent patterns of methylation (mainly high- or low-methylation). Maybe it could be something similar to finding peaks and valleys in ChIP-seq like tracks, I'm not sure.
For example, I have seen in some packages, such as ChAMP (R package), functions to define methylation blocks, but most are focused towards array data (which is sparse). Additionally, most of the standard DMR-finding packages have no "intermediate step" of definition of regions in single samples before they do comparisons.
Do you know of any software packages which can do this, or how do you usually tackle this problem?
Thanks for your help!
