Mitocondrial annotation irregularities

[passdan]

I know mitocondria annotation isn't supported and always been slightly irregular but I wanted to note it here incase any underlying cause can be found. Of a selection of 23 vertebrate mito genomes some annotate perfectly while some totally fail and I'm unable to see any difference causing this. Any suggestions would be appreciated. Will look at mitos etc in the future but this is for a taught course where the students will go on to bacterial annotation so a little mito annotation as a trial was always a nice entry point.

Straightforward loop over all assembled genomes. Most are complete, a few have at least 90% in one contig and ~10% in a second but that doesn't always correlate with the failing samples:

prokka --outdir prokka-${i} --kingdom Mitochondria --gcode 2 spades-${i}/contigs.fasta

### Successful (some mismatches but fine):

### Failed:

[andersgs]

@passdan is it related to the taxonomic group of the mitochondrial genome? some might have homologues that are closer to the models used in prokka than others. i suspect evolutionary distance might be playing a role.

if you are looking for a small genome to annotate as an example, you can try plasmids or just a small 10-15kb section of an E coli, Salmonella, or Staph genome.

mitoZ might be a more suitable tool for mito genomes if you are really keen on them: https://academic.oup.com/nar/article/47/11/e63/5377471

[passdan]

These are Bears and all pulled from Ensmbl so not too weird and should be pretty comfortably annotate-able in my opinion! Yes, for future might need to swap to bacterial.

That looks like a cool alternative, thanks!