Hello! I have some questions about the outputs from the

smrnaseq

pipeline which I can't seem to understand from the docs. First, I want to perform differential expression with the

differentialabundance

pipeline. I am not sure if I should be using the edger_qc/mature_counts.csv or the mirtop/mirna.tsv as input matrix. I understand that the former takes into account only the reads aligning perfectly to the mature RNA and the latter to all the different isomiRs that could be related to one mature miRNA. Is this correct? Another question, maybe more related to the experiment. The size selection was made to get RNA with size (aprox) < 65nt, so I do not expect to get pri-miRNAs other than the ones coming from degradation (?. But I get around 50% of my reads mapped to mature and 25% to hairpins. Is this expactable? Also, I plotted the size distribution of the sequences in mirtrace/qc_passed_reads.all.collapsed and got something like this. If the counts assigned to hairpin would be from degradation, shouldn't I expect a smoother line for the miRNAs, instead of the sharp peak at 22? Sorry for the silly questions, both miRNAs and nextflow are kind new for me 🙃

I think you should use mirtop quantification. You can as well, make your own matrix from mirtop files and isomiRs package if you want to ignore some files. But probably mirtop/mirna.tsv is ok.