BiSeq (version 1.12.0)

filterBySharedRegions: Reduces a BSraw or BSrel object to regions (or single CpGs) shared by a fraction of samples

Description

This method determines the regions which are covered in a given fraction of samples and reduces the BSraw or BSrel object to these regions.

Usage

filterBySharedRegions(object, groups, perc.samples, no.samples, minCov)

Arguments

object

A BSraw or BSrel object.

groups

OPTIONAL. A factor specifying two or more groups within the given object. See Details.

perc.samples

A numeric vector with elements between 0 and 1 of length 1 or of the same length as levels in group. Default is 1.

no.samples

Alternative to perc.samples. An integer vector of length 1 or of the same length as levels in group.

minCov

A numeric: If object is a BSraw object the minimum coverage may be set. Default is 1.

Value

object storing methylation information solely for regions or single CpGs covered in at least round(perc.samples*ncol(object)) samples, if perc.samples is given. Alternatively, the number of samples can be given directly by using no.samples.

Details

If argument group is given perc.samples or no.samples are applied for all group levels.

See Also

filterByCov

Examples

data(rrbs)
rrbs.reduced <- filterBySharedRegions(object = rrbs, perc.samples = 1)

# Or filter CpG sites where at least 50% of cancer and 50% of normal samples have coverage:
rrbs.reduced <- filterBySharedRegions(object=rrbs, groups = colData(rrbs)$group,
                                      perc.samples = 0.5)