limmaDE: Differential expression using limma
In beadarray: Quality assessment and low-level analysis for Illumina BeadArray data

Description Usage Arguments Details Value Author(s) Examples

Function to perform a standard limma analysis using a single command.

1	limmaDE(summaryData, SampleGroup, DesignMatrix = NULL, makeWts = TRUE, ...)

`summaryData`	ExpressionSetIllumina object
`SampleGroup`	Name of column in phenoData that will be used to construct sample groups for analysis
`DesignMatrix`	Optional design matrix
`makeWts`	if TRUE weights will be calculated for each array
`...`	Other arguments that lmFit can accept

The function automates the steps used in a typical limma analysis. Firstly, the lmFit is used to fit a linear model from the specified column in phenoData. Array weights can be calculated (arrayWeights) and used in the fit. A contrast matrix of all possible contrasts is created (makeContrasts) and fitted (makeContrasts). The empirical Bayes shrinkage of variances is then applied (eBayes). The design matrix, contrast matrix and array weights can all be extracted for the resulting object.

a limmaResults object

Mark Dunning

if(require(beadarrayExampleData)){
  data(exampleSummaryData)
  rawdata <- channel(exampleSummaryData, "G")
  normdata <- normaliseIllumina(rawdata)
  limmaResults <- limmaDE(normdata, SampleGroup = "SampleFac")
  limmaResults
  DesignMatrix(limmaResults)
}