adjustSignal: Array-based CGH Preprocessing

Description

This function performs several preprocessing steps: local regressions (loessFit) are used to correct cy3/cy5 and GC% bias, when Agilent dual-color hybridization are used only. In case of Affymetrix cychp.txt (or cnchp.txt) data are used, these steps have been already done in ChAS or Affymetrix Power Tools.

Usage

## S3 method for class 'rCGH':
adjustSignal(object, Scale=TRUE, Cy=TRUE, GC=TRUE, Ref="cy3",
    suppOutliers=TRUE, nCores=NULL, verbose=TRUE)

Arguments

object

: An object of class "rCGH"

Scale

: logical. If TRUE (default), Log2Ratios are standardized.

: logical. If TRUE (default), cy3/cy5 bias is corrected using a local regression (loessFit). For Agilent dual-color hybridization only. Notice that, in case of Affymetrix files (cychp.txt or cnchp.txt), this argument is automatically set to FALSE, since this step is managed when files are exported from ChAS or APT.

: logical. If TRUE (default), the GC% bias is corrected using a local regression (loessFit). For Agilent dual-color hybridization only. Notice that, in case of Affymetrix files (cychp.txt or cnchp.txt), this argument is automatically set to FALSE, since this step is managed when files are exported from ChAS or APT.

Ref

: string. Which cyanine was used as the reference. Possible values are "cy3" (default) and "cy5". For Agilent dual-color hybridization only.

suppOutliers

: logical. If TRUE (default), outliers are removed using an Expectation-Maximization algorithm (EM). See details below.

nCores

: numeric. The number of cores to use in order to speed up the computation. When NULL (default), half of the available cores are used. If a value greater than the number of available cores is passed, this latter will be used. See detectCores.

verbose

: logical. When TRUE (default), progress is printed.

Value

An object of class "rCGH"

Details

When suppOutliers is TRUE (default), the Log2Ratios are splitted with respect to chromosomes. The main regions within each chromosome are identified using EM. Within each region r_i, x[r_i] values are redifined according to the corresponding model parameters. as: $$x[r_i] ~ N(mu_i, sigma_i)$$ Notice that this step may substantially increase the computation time.

Examples

Run this code

filePath <- system.file("extdata", "Affy_cytoScan.cyhd.CN5.CNCHP.txt.bz2",
    package = "rCGH")
cgh <- readAffyCytoScan(filePath, sampleName = "AffyScHD")
cgh <- adjustSignal(cgh, nCores=1)
getParam(cgh)