Hi,

   I have 3 groups of samples(A,B,C), and for each group there are 6 samples(a1,a2,...a6; b1~b6; c1~c6), and I analyzed the data a couple of weeks ago and restored the DESeq2 subject dds to my harddisk. Later I also analyzed the cell type deconvolution and found that one sample (c3) is abnormal as the fraction of one cell type in obviously higher than those of other samples. So i want to remove c3 from dds, and re-analyze the data again. I could do it at the import step from kallisto import, but that data was uploaded to a remote server and I have to download them again. Is there anyway to do it from the current dds?

Thanks & Best regards,

Raymond

If idx is the number of the sample in colData(dds)

dds <- dds[,-idx]

DESeqDataSet is built on top of SummarizedExperiment, and all of these objects allow for matrix-like subsetting with either numeric, logical, or character vectors.

After removing samples you should rerun DESeq() to reestimate parameters.