jbrowse flails if it tries to parse a gff3 file with features of type 'gene' but no child features (like mRNA, etc). So it would be better to set the gff3 type in the ShortStack output to nc_RNA rather than the current gene, just for convenience when using jbrowse.
Also would be convenient to make a single gff3 (with DicerCalls of N and non-N combined), as well the current split ones.
Also in current version (3.2.1), there is a typo in the gff3 header. It reads "##gff-version3", but should read "##gff-version 3" on the first line of the gff3 files.
jbrowse flails if it tries to parse a gff3 file with features of type 'gene' but no child features (like mRNA, etc). So it would be better to set the gff3 type in the ShortStack output to nc_RNA rather than the current gene, just for convenience when using jbrowse.
Also would be convenient to make a single gff3 (with DicerCalls of N and non-N combined), as well the current split ones.