I am doing RNA-Seq data analysis on 22 samples from 3 batches. I have the RPKM values of all the samples. I performed PCA and I found some batch effect. I would like to remove the batch effect so that I can go ahead with differential expression analysis.
I was reading some posts and I saw that the removeBatchEffect() function in the limma package of bioconductor has been used a lot with count matrix. Can i use this function with a matrix containing RPKM values? In that case, will the result also be a matrix containing RPKM values with just the batch effects removed ? Also, should I take log2 values of the RPKM ?