Hi All, I know this is a very broad question and the field of miRNA is rapidly changing. I am looking for a solution (pipeline) that would help to analyze data comparing miRNA levels between 2 groups (20+ samples in each group, serum-derived miRNA, no replicates). Data is generated as PE150. As I understand (from previous work with bulk mRNA sequencing) there will be alignment portion and abundance/DE, plus potentially down-stream functional analyses. If there is one tool that is recommended and has an end-to-end solution between alignment/DE that would be great. I know that the alignment step is crucial - is bowtie1 the recommended tool? Which exact alignment settings should be used (seed length, how many mismatches, how many alignments per read allowed)? Should reads longer than "x" be removed? Is rRNA (or overall non-miRNA) depletion by alignment needed (where to acquire fasta sequences used by the miRNA community used to do that?)? Is the alignment done directly to whole-genome sequence (eg 38 or still hg19 is preferred?)? Which tool should be used to count individual, raw miRNA reads? Is Deseq2 still preferred over other tools (cufflinks, deseq1, edgeR)? Last but not least - can publicly available controls be used? I know that there are way too many kits available out there to prepare seq libraries and this might have a profound effect - would You recommend against using it at all?
I think you might find a previously asked question and its follow-up discussions helpful as a starting point: Question: miRNA differential expression