Analysing MIRA-seq methylation data

Question

[Saad Khan]

Hi,

I have a MIRA-seq data that I want to analyze. In the past I have used Bisulfite data for methylation analysis. There are a number of options available for Bisulfite data (e.g. aligners e.t.c. Aligning with Bismark etc and downstream analysis with methylkit/methpipe e.t.c.) I want to know what are the options or pipeline available for MIRA-seq data. Also what are the best practices for DNA methylation analysis with MIRA-seq.

Answer 1

[Charles Warden]

MIRA is a MeDIP-based strategy. In general, you'll want to use a normal aligner (such as BWA or Bowtie) and a ChIP-Seq/ChIP-chip peak-calling algorithm (MATS, MACS, etc.). I've used MATS on MIRA tiling array data, but I can't absolutely promise that is the best strategy.

[smk]

Can MEDIPS be used for the same purpose or do we have to use MACS/MATS?

Thanks

[Charles Warden]

I haven't tried it, but I Googled the MEDIPS Bioconductor package and I think it looks appropriate. There are lots of valid peak calling algorithms out there - I just think MATS/MACS are the most commonly used.

[smk]

How do you do differential methylation analysis once you do peak calling with MATS/MACS

[Charles Warden]

The peak calling compares the signal between the two groups (which perhaps has been normalized against an input reference). So, peaks called due to a significant difference in signal between the groups are most likely to be differentially methylated.