Dear all,

it´s probably a simple question, however I can´t find a proper solution yet. I process a 450k methlyation dataset with the minfi package. I noticed that there are 2523 NA values when I extract beta values from the RGChannelSet. The simple question is how I should proceed with these NA values i.e. how can I yield an RGChannelSet without features having NA?

Thanks,

Philipp

>RGSet

RGChannelSet (storageMode: lockedEnvironment)
assayData: 622399 features, 129 samples
  element names: Green, Red
An object of class 'AnnotatedDataFrame'
  sampleNames: 1 2 ... 129 (129 total)
  varLabels: ID txt ... param (106 total)
  varMetadata: labelDescription
Annotation
  array: IlluminaHumanMethylation450k
  annotation: ilmn.v1.2

>beta <- getBeta(RGSet)
>sumis.na(beta)) 
[1] 2523

Dear Kasper,

I use the latest version (minfi_1.16.1) & create the RGSet the standard way with read.450k.exp

baseDir <- system.file("extdata", package="minfiData")

targets <- read.450k.sheet(baseDir)
RGSet <- read.450k.exp(targets = targets)

My explanation would be that, since ß is the ratio between methylated probe intensity and total probe intensities (sum of methylated and unmethylated probe intensities), NA values can only occor if both methylated and unmethylated probe intensities are zero. Nevertheless, the question remains if this is worrisome and how I should proceed with these 2523 NA values in my RGSet.

Found similar (not too helpful) postings discussing missing values in the MethylSet  (minfi - missing values and minfi: Meth and Unmeth = 0 resulting in beta = NaN? )

Philipp