callBins: Call aberrations from segmented copy number data

Description

Call aberrations from segmented copy number data.

Usage

callBins(object, organism=c("human", "other"), method=c("CGHcall", "cutoff"), cutoffs=log2(c(deletion = 0.5, loss = 1.5, gain = 2.5, amplification = 10)/2), ...)

Arguments

object

An object of class QDNAseqCopyNumbers

organism

Either “human” or “other”, see manual page for CGHcall for more details. This is only used for chromosome arm information when “prior” is set to “all” or “auto” (and samplesize > 20). Ignored when method is not “CGHcall”.

method

Calling method to use. Options currently implemented are: “CGHcall” or “cutoff”.

cutoffs

When method=“cutoff”, a numeric vector of (log2-transformed) thresholds to use for calling. At least one positive and one negative value must be provided. The smallest positive value is used as the cutoff for calling gains, and the negative value closest to zero is used as the cutoff for losses. If a second positive value is provided, it is used as the cutoff for amplifications. And if a second negative value is provided, it is used as the cutoff for homozygous deletions.

...

Additional arguments passed to CGHcall.

Value

Returns an object of class QDNAseqCopyNumbers with calling results added.

Details

By default, chromosomal aberrations are called with CGHcall. It has been developed for the analysis of series of cancer samples, and uses a model that contains both gains and losses. If used on a single sample, or especially only on a subset of chromosomes, or especially on a single non-cancer sample, it may fail, but method “cutoff” can be used instead.

When using method “cutoff”, the default values assume a uniform cell population and correspond to thresholds of (assuming a diploid genome) 0.5, 1.5, 2.5, and 10 copies to distinguish between homozygous deletions, (hemizygous) losses, normal copy number, gains, and amplifications, respectively. When using with cancer samples, these values might require adjustments to account for tumor cell percentage.

Examples

Run this code

data(LGG150)
readCounts <- LGG150
readCountsFiltered <- applyFilters(readCounts)
readCountsFiltered <- estimateCorrection(readCountsFiltered)
copyNumbers <- correctBins(readCountsFiltered)
copyNumbersNormalized <- normalizeBins(copyNumbers)
copyNumbersSmooth <- smoothOutlierBins(copyNumbersNormalized)
copyNumbersSegmented <- segmentBins(copyNumbersSmooth)
copyNumbersSegmented <- normalizeSegmentedBins(copyNumbersSegmented)
copyNumbersCalled <- callBins(copyNumbersSegmented)