| coerce {convert} | R Documentation |
Convert between limma, marray and Biobase data objects.
Objects can be converted (coerced) from one class to another using
as(object, Class) where object is an object to convert
and Class is the name of the class to convert to. The following
conversions are provided:
| From: | To: |
RGList | marrayRaw |
marrayRaw | RGList |
MAList | marrayNorm |
marrayNorm | MAList |
RGList | NChannelSet |
marrayRaw | NChannelSet |
MAList | ExpressionSet |
marrayNorm | ExpressionSet |
| Deprecated | |
RGList | exprSet |
MAList | exprSet |
marrayRaw | exprSet |
RGList and marrayRaw are coerced to
NChannelSet. Channel values are not transformed.
MAList and marrayNorm are coerced so that the
ExpressionSet slot contains log-ratios (M-values) and the
ExpressionSet object has the same number of columns as the
original object. In this case, information on the A-values is lost.
There is intentionally no conversion from RGList or
marrayRaw to ExpressionSet, as ExpressionSet is
intended for expression values, not intensities.
exprSet objects are deprecated, but converters to this class
exist. RGList and marrayRaw objects are coerced to
exprSets in such a way that green and red channel intensities are
stored in alternate columns of the exprs slot expression matrix.
This means that the resulting object has twice as many columns as the
original object.
Gordon Smyth and others
as in the methods package.
##first set up some fake intensity matrices
testRed = matrix(rnorm(5*2),5,2,
dimnames=list(paste("gene",1:5, sep=""), c("S1", "S2")))
testGreen = matrix(rnorm(5*2),5,2,
dimnames=list(paste("gene",1:5, sep=""), c("S1", "S2")))
##some sample/target info
testTarget = data.frame(slide=c("S1", "S2"), Cy3=c("T", "C"),
Cy5=c("C", "T"), row.names=c("S1", "S2"))
maT = new("marrayInfo", maLabels=c("S1", "S2"),
maInfo= testTarget)
##now create instances and convert
x <- new("RGList")
x$R = testRed
x$G = testGreen
y <- as(x,"marrayRaw")
z <- as(x, "NChannelSet")
x <- new("marrayRaw")
x@maGf <- testGreen
x@maRf <- testRed
x@maTargets = maT
y <- as(x,"RGList")
z <- as(x, "NChannelSet")
x <- new("MAList")
y <- as(x,"marrayNorm")
##we construct a reasonably complete fake, small
##instance of the marrayNorm class
x <- new("marrayNorm")
x@maM <- testRed
x@maA <- testGreen
maTargets(x) = maT
y <- as(x,"MAList")
y <- as(x, "ExpressionSet")
x <- new("MAList")
x$M <- testRed
x$A <- testGreen
x$targets = testTarget
y <- as(x,"ExpressionSet")