normalize.AffyBatch.vsn {vsn} | R Documentation |
Wrapper for vsn
to be used as a normalization method in the package affy
normalize.AffyBatch.vsn(abatch, subsample=20000, niter = 4, ...)
abatch |
An object of type AffyBatch . |
subsample |
The number of probes to be sampled for the fit of the transformation parameters. |
niter |
Parameter passed on to vsn . |
... |
Further parameters for vsn . |
Please refer to the "details" and "references" sections of the man page for
vsn
for more details about this method.
Important note: after calling vsn
, the function
normalize.AffyBatch.vsn
exponentiates the data. This is done in
order to make the behavior of this function similar to the other
normalization methods in affy. There, it is assumed that in subsequent
analysis steps (e.g. in medianpolish
), the logarithm to
base 2 needs to be taken.
An object of class AffyBatch
.
Wolfgang Huber mailto:w.huber@dkfz.de
library(affy) library(affydata) data(Dilution) ## let affy know about vsn normalize.AffyBatch.methods <- c(normalize.AffyBatch.methods, "vsn") es1 = expresso(Dilution[1:2], bg.correct = FALSE, ## bg correction is done by vsn normalize.method = "vsn", pmcorrect.method = "pmonly", summary.method = "medianpolish") es2 = expresso(Dilution[1:2], bgcorrect.method = "rma", normalize.method = "quantiles", pmcorrect.method = "pmonly", summary.method = "medianpolish") ## graphics output if(interactive()) x11() oldpar = par(mfrow=c(2,2), pch=".") ## extract expression values x1 = exprs(es1) x2 = exprs(es2) ## scatter plot plot(x1, main="vsn: chip 3 vs 4") plot(x2, main="rma: chip 3 vs 4") ## rank(mean) - difference plot ylim = c(-0.7, 0.7) plot(rank(rowSums(x1)), diff(t(x1)), ylim=ylim, main="rank(mean) vs differences") abline(h=0, col="red") plot(rank(rowSums(x2)), diff(t(x2)), ylim=ylim, main="rank(mean) vs differences") abline(h=0, col="red") ## reset old plotting parameters par(oldpar)