RPKM/CPM/FPKM to log2 units

Question

[zhang.jianhai]

Hi,

In network analysis on RNA seq data. Should the RPKM/CPM/FPKM be converted to log2 units?

Regards

Answer 1

[dsull]

What type of network analysis are you doing?

In general, it's best to use counts normalized in DESeq2 via the varianceStabilizingTransformation function.

For example, if you're doing WGCNA, the FAQ suggests the following:

"We then recommend a variance-stabilizing transformation. For example, package DESeq2 implements the function varianceStabilizingTransformation which we have found useful, but one could also start with normalized counts (or RPKM/FPKM data) and log-transform them using log2(x+1). For highly expressed features, the differences between full variance stabilization and a simple log transformation are small." - https://horvath.genetics.ucla.edu/html/CoexpressionNetwork/Rpackages/WGCNA/faq.html

[zhang.jianhai]

I am doing gene co-expression network analysis on RNA-seq data. The only expression matrix I received is RPKM/FPKM/CPM, so I am wondering if it should be further log2 transformed?

Regards.

[dsull]

If you're going to use FPKM, then yes, you should log2-transform (which is exactly what the FAQ above states so just go off of that).

[zhang.jianhai]

Got it. Thank you so much!