Remove outlier from ExpressionFeatureSet - Outliers

Question

[ab123]

Hello there,

Does anybody know how to remove a dataset from an ExpressionFeatureSet? I've plotted quite a few graphs and checked outliers and one of the CEL datasets came back as an outlier, but I cannot seem to find any information on how to omit it from loading.

There must be a way e.g. to load it with list.celfiles and somehow omit one of the datasets from loading?

Or alternatively delete it once loaded from the ExpressionFeatureSet. I've been using oligo.

I'd be grateful for any advice.

PS. In terms of outliers, there seem to be very conflicting views on when to remove it. Since I cannot be sure about any technical issues since I'm using existing data, I'm not sure exactly when to delete one. If anybody has any thoughts on that please feel free to share. E.g. do you make a difference between small/large data sets.

Cheers

Answer 1

[theobroma22]

If you download CEL files to your directory you can load whichever you want in R using the oligo package.

Answer 2

[h.mon]

As you do not say how you loaded the files, I will give you three alternatives, which may or may not apply to your case:

1) move the offending CEL file to a different place before reading the files into R

2) remove the file from the list of files you will read (assuming the third file is the outliar):

ff <- list.files( path = ".", pattern = "*.cel$", full.names = TRUE )
ff <- ff[ -3 ]

3) remove the data from the ExpressionSet objet:

eset <- eset[ , -3 ]