Hi. When

--save_reference

is on, under the

genome

dir, there is a file

*.filtered.bed

. It was generated by

gtf2bed [filtered.gtf] > [filtered.bed]

. I noticed that it only has ~50 lines while there are more than 2000 in filtered.gtf. I have a few questions about this: • Only these ~50 lines are kept because they received an "exon" label in the input filtered.gtf. The gtf2bed code effectively selects only "exon" lines implemented by the line 54 block. This block only keeps exon, start/end_codon, and miRNA. Is this an intended behavior for bacterial GTF which has majority of the records as CDS? • Since

exon

does not exist in my input GTF, it must get added by a previous step (presumably gtf_filter.py though a simply search didn't find a line doing that). I noticed that only the rRNA and tRNA records in my input gtf received an exon label in the 3rd column in filtered.gtf, while all the remaining keep the CDS label. Again, I wonder if this is an intended behavior. • I didn't check the workflow deeply, but I'm not seeing this "filtered.bed" file is used by any subsequent steps. If it's really unused for anything when

-profile prokaryotic

is on, it's OK to have it there (though might be a little confusing). If it's used for something, I wonder what's the reason to limit to rRNA and tRNA. Thanks!

Hi! I fixed something just today, see if the release below helps. I basically moved the prokaryotic path to bypass gtf2bed. In general, could you start raising your issues / bug reports on GitHub please? This channel is for more general questions / support and you've a much better chance of me addressing something if you've added it to github with enough information for me to replicate a problem. Thank you!

Sure! Sorry about that. I recently had experience of being directed from github issues to slack or other platforms. Trying to remember the best manner at different places 🙂

Technical ask = GitHub (which most of your excellent points have been). Also allows me to credit you better. General question/ discussion = Slack