identifySigDMR: Identify significantly DMR (Differentially Methylated Region)

Description

Identify significantly DMR (Differentially Methylated Region)

Usage

identifySigDMR(detectResult, p.adjust.method = "fdr", pValueTh = 0.01, fdrTh = pValueTh, diffTh = 1, minProbeNum = 1, maxGap = 2000, minGap=100, oppositeMethylation = FALSE, topNum = NULL)

Arguments

detectResult

A GRanges object or a data.frame with "PROBEID", "CHROMOSOME" and "POSITION" columns

p.adjust.method

p.value FDR adjust method

pValueTh

The threshold of p.value

fdrTh

The threshold of FDR (adjusted p.value)

diffTh

The threshold of difference between two conditions

minProbeNum

Minimum number of probes in each DMR

maxGap

The maximum gap allowed between two nearby probes to be considered within a same DMR

minGap

If two nearby DMRs have a gap less than or equal to the minGap, they will be merged as a single DMR

oppositeMethylation

Whether require the averaged methylation levels in the DMR are in opposite direction

topNum

Whether only returns the top number of probes (ranked by p.value)

Value

A list of GRanges objects, sigDMRInfo and sigDataInfo. sigDMRInfo contains DMR information, while sigDataInfo includes the probe level information within the DMRs.

Details

(with "status" column)

We define a differentially methylated region (DMR) as a region, in which most measured CpG-sites are differentially methylated. To identify DMRs, the function first determines the differential methylation status of each probe, then merge them as a continuous region. The identifySigDMR function returns a list of two GRanges objects. The sigDMRInfo includes the identified DMRs, and the sigDataInfo includes all differentially methylated probe information.

Examples

Run this code

	data(exampleMethyGenoSet)
	
	sampleType <- colData(exampleMethyGenoSet)$SampleType
	
	## Do differential test
	allResult <- detectDMR.slideWin(exampleMethyGenoSet, sampleType=sampleType, testMethod='ttest')

	## Identify the DMR (Differentially Methylated Region) by setting proper parameters.
	## Here we just use default ones
	allDMRInfo <- identifySigDMR(allResult)