Which package to use for RPKM/FPKM RNA-seq counts?

Question

[hafiz.talhamalik]

I have RNA-Seq counts which are FPKM/RPKM normalized. which package in R should i use for Differential gene expression. among many packages which package is better? is DGE using raw count is better or using FPKM/RPKM normalized counst are better ?

Thanks in anticipation.

Answer 1

[Matteo Schiavinato]

which package in R should i use for Differential gene expression?

Assuming you mean to do diff.expr. with FPKM, I think Limma does a pretty good job. However ...

among many packages which package is better?

This question I think is hard to answer. The jury is out on this one!

is DGE using raw count is better or using FPKM/RPKM normalized counst are better ?

The author of the FPKM measurements himself called FPKM as "biased" and "wrong". So I would go for raw counts when possible. My pipeline usually includes htseq-count and DESeq2, but other couting programs and other differential expression analysis programs are fine!

[hafiz.talhamalik]

Thanks.. but limma was built for micro-array is it good enough for RNA-seq ?

[Matteo Schiavinato]

As far as I know, you can plug FPKM values in.

Answer 2

[Renesh]

DESeq2 is a good package for DEG analysis in R. https://bioconductor.org/packages/release/bioc/html/DESeq2.html

If you are using R package for DEG analysis, then go ahead with raw counts. It is not recommended to perform DEG analysis on normalized count using DESeq2 because it performs internal normalization.

[hafiz.talhamalik]

Thankyou so much.. and what about edgeR ?