using liftoff with bacterial genomes?

Question

using liftoff with bacterial genomes?

Closed this issue 3 years ago · 2 comments

I was wondering whether there would be any issues in applying liftoff to bacterial genome annotations. I found one issue in https://github.com/agshumate/Liftoff/blob/master/liftoff/polish.py where something like protein = longest_ORF.translate() is run using the default translation table. I'm wondering whether it might be better to identify the translation table in the reference annotation (like from the transl_table qualifier) and then use that everywhere. Or making it a flag.

Answer 1 · 2021-12-03T16:02:45.000Z

Hi,
the -polish option does use the default translation table. . However since bacterial genome annotations are much simpler, you should be able to get very good results without needing the -polish option

Answer 2 · 2021-12-08T05:36:03.000Z

Oh, we already had a script to find reference genes by blasting to our sample genome. We thought it'd be better to instead use liftoff to get extra QA steps like this--making sure beginnings and ends are start/stop codons, etc.