Methylation analysis toolkit
For two samples, create a bed file with rows representing CpG Locations, and two columns -- one fore each sample.
Then, run fisherImplementedIndCpg.pl
to run Fisher's exact test on each CpG. This adds a column for the p-value and the absolute difference in methylation.
Next, run defineRegions.pl
to select DMRs using a simple sliding window approach.
The paramters for defineRegions.pl
are:
$file
= file generated byfisherImplementedIndCpg.pl
$diffCutoff
= minimum methylation difference for a CpG to be considered significant, default = 0.5$pvalCutoff
= minimum p-value for a CpG to be considered significant, default = 0.01$minSigCount
= minimum number of significant CpGs in a window for it to be added to the final list of regions, default = 3$windowSize
= size of sliding window, default = 5$resetDistance
= default = 1000
Thus, with the defaults, windows where at least 3/5 CpGs (with gaps of no greater than 1000bp between the CpGs) with a methylation difference of at least 0.5 and p-value < 0.01 will be called significant.