I ran conditional quantile normalization on my RNAseq counts. I am then going to use DEseq2. However, DEseq2 only accepts non-negative integers. CQN can be coerced to be non-negative, but is still on the log scale and cannot be coerced into integers. So because DESeq2 only accepts un normalized non-negative integers, then how can the merits of CQN be considered when calculating for differentially expressed genes using DESeq2? Either CQN needs to have an integer normalized output or DESeq2 needs to accept non-integers.
Any help on this is appreciated.