Differentially methylated regions analysis

[Jluis]

Hello everybody,

I'm trying to perform a Methyl-seq analysis where the aim is to obtain differentially methylated regions (DMRs) between two groups of individuals (mice) grown in different conditions. Nine individuals were sampled on each group and I want to compare their methylation profiles.
I usually perform Methyl-seq analysis using Bismark to map reads and do the methylation extraction and then I usually display those data using Seqmonk. In this particular case I need to find a tool that allows me to perform a DMR analysis with all those samples. I read about MEDIPS, but I don't really understand if this tool cam compare more than one sample for each condition in the study or in the case of having more than one sample for each condition the tutorial states:
"Region informations from several mapping results have to be pooled into one file." (I may be wrong but after reading the MEDIPS tutorial, it seems to me that you have to pool your replicates for each condition to perform the analysis).

Sumarizing, my question is:
Do any of you have any suggestion or knowledge about any software that can accurately perform DMR analysis using as input multiple biological replicates for each condition studied?

Thanks in advance

[cwhelan]

I can't speak to their accuracy yet but you might want to look into two new R packages for methylation data:

bsseq:

http://www.bioconductor.org/packages/2.11/bioc/html/bsseq.html

methylkit:

http://code.google.com/p/methylkit/

I am not sure how they would work with Methyl-seq but they might be a starting point for you to investigate.

[Jluis]

@cwhelan

Thank you very much for your advice, I'll give a try to those modules and see if are suitable for the tasks I need to perform.

With best wishes

[Jluis]

@cwhelan - I presume you're familiar with methylkit. If I'm right, May you know where to obtain the bed format annotation files stated in the methylkit manual?:

"# read-in transcript locations to be used in annotation
# IMPORTANT: annotation files that come with the package (version >=0.5) are a subset of full annotation files.
# Download appropriate annotation files from UCSC (or other sources) in BED format"

Thanks in advance

[cwhelan]

@Jluis

I am not terribly familiar with methylkit; I was just reading about it because I've been writing a similar set of scripts for analysis of WGBS data and ran across it doing my research.

For the annotations, check the next section of their manual, "Downloading Annotation Files":

https://code.google.com/p/methylkit/#Downloading_Annotation_Files

[Jluis]

@cwhelan

Thank you for the advice, seems I was reading a manual where that info wasn't specified.
Do you plan to release your WGBS scripts after publication? It would be nice to have that resource available for this kind of analysis in the near future.

Best wishes

[cwhelan]

If I end up with anything useful that offers something that these new packages don't I'll be sure to try to make that code available.

[xfliwz]

Hello Jluis,
Our group have developed a sliding window approach to identify DMR called swDMR.
Detail information can be get at http://seqanswers.com/forums/showthread.php?t=23196.
Hope this may help you.

Yours truly
xfliwz

[Jluis]

@xfliwz

Sorry for the delay answering... but anyway thank you for the advice, I'll take a look at this tool and see if I can use it for my study.

Best wishes