We have very nice functions like DESeq2::rlog(), DESeq2:: varianceStabilizingTransformation() or DESeq2::vst() to get abundance estimate from count data which is much more usable/reliable than ad hoc transformations like log2(CPM+1).

For readers new to this topic note that the author/maintainer of DESeq2 recomends the vst() approach according to this post and that DESeq2::vst() is a fast wrapper for DESeq2:: varianceStabilizingTransformation().

My question is: whether something similar can be done to abundance values (CPM, RPKM, FPKM, TPM etc) for cases were we cannot get the underlying count data?