SeqExpressionSet can store unnormalised counts and normalised counts, but it only seems to be possible to plot the unnormalised counts by
meanVarPlot. It would be useful to be able to plot normalised counts to visualise the effect of applying a normalisation method (e.g. DESeq2's
varianceStabilizingTransformation) to the count data set (i.e. is it visibly homoscedastic once it's transformed?). Also, if a
SeqExpressionSet is created by
newSeqExpressionSet, there is a table of
NA created in
normalizedCounts. If a table of normalised values is added later by using
normCounts(set) <- measurementsVST, then an element named
normCounts is added to
set@assayData. So there are two tables for normalised values simultaneously in
assayData. Can this inconsistency be resolved?
This is a bug and should be fixed. I'll take care of this ASAP.
normCounts(set) <- measurementsVST should store the normalized counts in the
meanVarPlot function uses the normalized counts, if available. But of course it looks for it as
normCounts. So fixing the bug will also make the behavior of the function what you expect.