I'm currently trying to run miRDeep2 in order to find, extract and count miRNAs from a miRNA sequencing experiment, but I'm having trouble determining which species to use as 'close/related species'.
My main question is: how many species should I use as 'close/related species'? Would using a big number of close species be better for the final output, or worse? And how closely related should the species be to my reference?
I'm working with sequences from Bos indicus, using the genome and miRNA from Bos taurus as reference, and when asking other people in the lab, I've got completely opposite directions and pipelines, which ended up being very misleading.
- Use only one species as 'close species', using Homo sapiens, as it's a "very well studied species".
- Use more species, in order to have more sequences and find more miRNAs, using Equus caballus, Mus musculus, Homo sapiens, Sus scrofa and the sequences from Bos taurus itself as close species (5 species).
- Use more closely related species, in this case, Ovis aries and Capra hircus, both ruminants.
The first two came from people who already worked (or are working on) similar analysis as my own, although I'm pretty sure using the same species as both reference AND close species would be wrong, even if the data we have is from Bos indicus, not Bos taurus.
The idea of using either Ovis aries or Capra hircus as reference seems appealing to me, as they are more closely related to Bos taurus than the other mentioned ones.
At the moment, I'm considering running it with Ovis aries, Capra hircus and Homo sapiens, but I'm still unsure about how closely related species should be, and how many I should use.