I have a question about using voom() on TPM (or RPKM) values. I have TPM values on gene level for 200 samples, and I want to use voom() function to log-normalize the data. I will use the output voom(dge,design,plot=FALSE)$E, from Voom, to run a linear regression model myself for DE analysis (so I'm NOT using Limma afterwards).
My question is whether this is a correct thing to do? In other words, is the log-normalization method used in voom() applicable to RPKM values?
If this is not a correct thing to do, what other method you suggest for normalization or transformation of the RPKM values before doing a DE analysis?
I'd deeply appreciate any kind of comment or suggestion.
Thanks,
Golsheed
My question is whether this is a correct thing to do? In other words, is the log-normalization method used in voom() applicable to RPKM values?
If this is not a correct thing to do, what other method you suggest for normalization or transformation of the RPKM values before doing a DE analysis?
I'd deeply appreciate any kind of comment or suggestion.
Thanks,
Golsheed