splitread todo
to do
1/ add filtering logic of the candidates (intron location)
- use a filtering flag (explain why candidates have not been retained) instead of a binary value
- add filtering depth threshold (command line and in file, default = 3)
- add filtering on candidate length (over 80% of the contig length)
- add filtering on canonical intron boundaries (on both strand)
- less than non overlapping 3 candidates