Normalize bulk-rna seq with median of ratios and DEseq2

Question

[anael.cain]

Hi,

I have multiple samples from bulk-rna seq that I want to normalize using the median of ratios method, using DEseq2. I probably make something wrong as the result that I get before and after the normalization are pratically similar to what I get if I make a FPKM score. I thought that calculating the sizeFactor before the normalization step is enough to perform the median of ratios. What I'm missing? Here is the code that I'm using - all the samples should be normalized together

dds <- DESeq2::DESeqDataSetFromMatrix(countData = counts,
                                      colData = infos.samples, design = ~1)
dds <- estimateSizeFactors(dds)
counts.norm <- t(DESeq2::counts(dds, normalized = T))

In addition, is there a way (maybe using the design) to normalize across a condition?

Thank you!

[ATpoint]

Do you need a design? No, the design has nothing to do with how the method works.

If you have no compositional changes between your samples and only the depth of sequencing is the difference between the samples then FPKM and DESeq2 should yield fairly similar outputs. What exactly does surprise you?

[anael.cain]

You're right. I'm new to this and thought that different methods yield different results.

Thank you for the help