transformed read counts in deseq2

Question

[anna]

How can I extract the normalized and transformed read counts in deseq2 which are used to calculate the log fold change?

[GenoMax]

See answer here: retrieve normalised count data from DESeq2

[anna]

i've tried that function but the data are not transformed. Also i've read that the function (rlog) retrieve the transformed counts but i can't get any result by doing this: write.csv(as.data.frame(rlog(dds, blind=TRUE, fitType = "parametric")), file="transformed.csv")

[swbarnes2]

What function exactly did you try? rlog values are not used for calculating log fold changes.

Answer 1

[ATpoint]

rld <- rlog(dds)
assay(rld)

How to export rld (rlog) or vsd values from DESeq2?

As far as I know essentially all DESeq2 object types are SummarizedExperiment containers and therefore can be accessed with assay to select counts. data.frame(rlog) is not possible as you first have to access the counts inside the SummarizedExperiments container.

Answer 2

[Prakash]

you can also extract normalized read count from dds object

res <- as.data.frame(counts(dds, normalized=TRUE))