decide on miRNA data normalization

[MvanderGragt]

Does the abundance of mature miRNAs equal MIR gene expression? Are methods used for normalization of Protein-Coding genes suitable for miRNA normalization? -> the mapped reads show a different pattern then with P-C genes.

[mgalland]

The abundance of the mature miRNA ("MajorRNA" in Shortstack terminology) does not equal the MIR gene expression since the MIR gene length is longer than the mature miRNA. More sRNA-Seq reads will be counted on the MIR gene.
Still, the mature miRNA count values should be strongly correlated to that of the MIR gene. Perhaps something to plot?

As for your second question on protein coding genes, I'd say that, while the read mapping pattern is different, the statistical normalisation method should be comparable since both read counts should follow the negative binomial distribution. For additional details, please refer to the DESeq method paper: https://genomebiology.biomedcentral.com/articles/10.1186/s13059-014-0550-8