Differential expression analysis on normalized RNAseq dataset

Question

[BlueSky]

I have a ready-to-use normalised (already harmonised) RNAseq dataset, and I want to find differentially expressed genes from the dataset. In the article it says that the "RNAseq expression data were normalised using the upper quantile method implemented in the R package NOISeq v2..31.0".

In the paper it also says in the "further usage" section that differential expression analysis on the RNAseq dataset can be done by DESeq2, NOISeq or edgeR. What confuses me is that when I look at the description for DESeq2, they want an un-normalized expression matrix. Can I still use DESeq2 on the normalised dataset as it says in the paper or will that be a problem? How do I solve it?

Answer 1

[ATpoint]

This thread summarizes what you can do from normalized counts:

https://support.bioconductor.org/p/125144/#125156

Please also use google and the search function, it is one of the top FAQs.