Entering edit mode
Hi,
I was wondering how should one go about normalising 450K Methylation data from GEO that has unmethylated and methylated signal (the file name state that it is unprocessed). I have tried using readGEORawFile() in minfi and the output object is a GenomicMethylSet, which unfortunately does not meet the requirement of the normalisation methods which requires RGChannelSet as the input.
From my understanding it is NOT possible to convert MethylSet to RGChannelSet (e.g. as certain info such as control probes are missing) I would like to know how does the bioconductor community handle this situation.
Thanks!
If you have a MethylSet you will still be able to use quantile normalisation methods that are described in the wateRmelon package (I recommend dasen for general purposes). However the currently functionality of wateRmelon with minfi isn't a nice as it could be but this will be addressed in Bioconductor 3.5.
The GEO accession may have the raw data files (.idats) located in another place - which would allow you to import the data as an RGChannelSet and thus be able to use a minfi normalisation method.