[BioC] Error normalizing merged data sets

Amit Kumar amitkumartiwary at gmail.com
Tue Sep 29 22:33:23 CEST 2009

Dear All,

I am working with 13 different data sets (containing 550 arrays in
total) which is normalized using general linear model for print-tip. I
am trying to do following:

1) Merging all 13 different data sets:
To merge these data sets I built ExpressionSet for each data set. Then
I made a list of these ExpressionSets and named it "fullset1". Then to
merge red and green intensity I did following:

red1 = lapply(fullset1,function(eset) exprs(eset)[2:23689,seq(1,
ncol(eset), 4)]) # These columns contain normalized Red intensity in
each data file.
red = do.call(cbind, red1)

green1 = lapply(fullset1,function(eset) exprs(eset)[2:23689,seq(2,
ncol(eset), 4)]) # These columns contain normalized Green intensity in
each data file.
green = do.call(cbind, green1)

2) Normalizing between arrays:
I am using normalizeBetweenArrays (method=quantile) for merged data
set so that the intensities have similar distributions across the
series of arrays. So I did following:

rgl1 = new("RGList", list(R=red, G=green))
norm1 = normalizeBetweenArrays(rgl1, method="quantile")

And got error: Error in log(R, 2) : Non-numeric argument to
mathematical function. Why this error??

Then I tried to normalize red and green intensities separately by doing this
normR1 = normalizeBetweenArrays(red, method="quantile")
normG1 = normalizeBetweenArrays(green, method="quantile")
and it worked! Why??

3) Quality assessment of the merged data set:
For quality assessment I am using arrayQualityMetrics package. To
check if quality assessment on separately normalized red and green
intensities can be performed I did following:
rgl1 = new("RGList", list(R=normRed1, G=normGreen1))
qmrgl1 = arrayQualityMetrics(rgl1, "QMrgl1") # Here the package is
accepting expression data as RGList (rgl1) and will generate HTML
report and save in QMrgl1 folder.

But with no luck I got error: Error in rc - gc : non-numeric argument
to binary operator

I am really stuck and running out of idea. Please help me to rectify
these errors. Thanks in advance.


More information about the Bioconductor mailing list