Perform ARRm normalization

Share:

Description

This function perform Adaptive Robust Regression method (ARRm) normalization on Beta values. The method corrects for background intensity, dye bias and spatial on-chip position. By default, chip mean correction is also performed.

Usage

1
2
normalizeARRm(betaMatrix, designInfo, backgroundInfo, outliers.perc = 0.02, 
goodProbes = NULL,chipCorrection=TRUE)

Arguments

betaMatrix

matrix containing the Beta values. Rows are probes, columns are samples.

designInfo

A data.frame containing a column named chipInfo containing the chip indices and a column named positionInfo containing the position indices

backgroundInfo

A data.frame containing two columns: green contains the background intensity in the green channel for each sample and red contains the background intensity in the red channel for each sample

outliers.perc

Proportion (between 0 and 1) of outliers to be removed from the ARRm regression

goodProbes

Ids of the probes to be normalized (Id. of the form "cg00000029")

chipCorrection

logical, should normalization correct for chip mean?

Value

A matrix containing the normalized Beta values

Author(s)

Jean-Philippe Fortin <jfortin@jhsph.edu>

See Also

getBackground to see how to obtain background information from control probes, and getDesignInfo to see how to obtain position and chip indices

Examples

1
2
3
4
5
6
7
data(greenControlMatrix)
data(redControlMatrix)
data(sampleNames)
data(betaMatrix)
backgroundInfo=getBackground(greenControlMatrix, redControlMatrix)
designInfo=getDesignInfo(sampleNames)
normMatrix=normalizeARRm(betaMatrix, designInfo, backgroundInfo, outliers.perc = 0.02)