correct order of cut off and normalization in RNAseq data

Question

[Sara]

I am a bit confused about applying cut off and normalization. for my RNAseq data I have raw counts, cpm and FPKM. I have 2 Scenarios to apply cut off:

1- using cpm: applying cut off and then normalized the raw counts (the normalized data would be cpm)
2- using FPKM: normalizing the raw count (the result would be FPKM) and then applying cut off.

do you guys know if I am using the correct order (cut off and normalization) in each scenario?

Answer 1

[dsull]

This is not how you do RNA-seq.

You must feed your raw counts into a RNA-seq package like DESeq2 -- it normalizes things the correct way and then you will then be able to set your cutoffs (based on log2 Fold Changes and adjusted p-values).

Also, in the future, please describe your experiment in more detail: e.g. "I have six samples: three treated samples and three control samples, and would like to identify genes that are differentially expressed between the treated and control groups"