runCBS: Run CBS Segmentation

In genoset: A RangedSummarizedExperiment with methods for copy number analysis

Description

Utility function to run CBS's three functions on one or more samples

Usage

runCBS(data, locs, return.segs = FALSE, n.cores = 1,
  smooth.region = 2, outlier.SD.scale = 4, smooth.SD.scale = 2,
  trim = 0.025, alpha = 0.001)

Arguments

data	numeric matrix with continuous data in one or more columns
locs	GenomicRanges, like rowRanges slot of GenoSet
return.segs	logical, if true list of segment data.frames return, otherwise a DataFrame of Rle vectors. One Rle per sample.
n.cores	numeric, number of cores to ask mclapply to use
smooth.region	number of positions to left and right of individual positions to consider when smoothing single point outliers
outlier.SD.scale	number of SD single points must exceed smooth.region to be considered an outlier
smooth.SD.scale	floor used to reset single point outliers
trim	fraction of sample to smooth
alpha	pvalue cutoff for calling a breakpoint

Details

Takes care of running CBS segmentation on one or more samples. Makes appropriate input, smooths outliers, and segment

Value

data frame of segments from CBS

See Also

Other 'segmented data': bounds2Rle, rangeSegMeanLength, segPairTable, segTable, segs2Granges, segs2RleDataFrame, segs2Rle

Examples

    sample.names = paste('a',1:2,sep='')
    probe.names =  paste('p',1:30,sep='')
    ds = matrix(c(c(rep(5,20),rep(3,10)),c(rep(2,10),rep(7,10),rep(9,10))),ncol=2,dimnames=list(probe.names,sample.names))
    locs = GRanges(ranges=IRanges(start=c(1:20,1:10),width=1,names=probe.names),seqnames=paste('chr',c(rep(1,20),rep(2,10)),sep=''))

    seg.rle.result = RleDataFrame( a1 = Rle(c(rep(5,20),rep(3,10))), a2 = Rle(c(rep(2,10),rep(7,10),rep(9,10))), row.names=probe.names )
    seg.list.result = list(
      a1 = data.frame( ID=rep('a1',2), chrom=factor(c('chr1','chr2')), loc.start=c(1,1), loc.end=c(20,10), num.mark=c(20,10), seg.mean=c(5,3), stringsAsFactors=FALSE),
      a2 = data.frame( ID=rep('a2',3), chrom=factor(c('chr1','chr1','chr2')), loc.start=c(1,11,1), loc.end=c(10,20,10), num.mark=c(10,10,10), seg.mean=c(2,7,9), stringsAsFactors=FALSE)
      )

    runCBS(ds,locs)  # Should give seg.rle.result
    runCBS(ds,locs,return.segs=TRUE) # Should give seg.list.result

genoset documentation built on Nov. 8, 2020, 6:07 p.m.