DESeq2 log2 fold change

Question

[bioinfonerd]

Hi everyone,

I did the DESeq2 on my Nanostring gene expression data using the below code:

dds <- DESeqDataSetFromMatrix(countData = counts_set,colData = annotation,design = ~ pt_id+treatment)
sizeFactors(dds)  <- estimateSizeFactorsForMatrix(counts(dds))
dds <- DESeq(dds)
res <- results(dds,contrast=c("treatment","Treatment","Vehicle"))

But the log2FC I am getting are quite different from the log2FC I calculated manually. Could you let me know if that is normal and correct?

Thanks!

[h.mon]

Every now and then this question appears here:

Discrepancy in DESeq2 fold change

How to recover treated/control count from DESeq2 output

Why can't I replicate DESeq's Log2FC calculation?

[ATpoint]

I guess that your FCs for the genes with smaller couts are much larger than the DESeq2-ones?

EDIT: You should have a look at this post from the DESeq2 developer towards DESeq2 on Nanostring data. He is concerned that the normalization might be a problem.