martinghunt

I can try to debug. Could you please provide example data, the command(s) run to reproduce the output, and version of ariba you're using?

Thanks, yes we know that this can happen. I could add in an option to not require a full length match to the reference gene? I'm thinking require the start...

Could you copy + paste the contents of `/home/zong/Desktop/666/96200/circlator/spades/04.merge.merge.iter.2.assembly/contigs.paths` here please?

When you run circlator all, it tries to circularize all the contigs. The names of the ones that could not be circularized are written to the file `06.fixstart.contigs_to_not_change`. When fixstart...

Maybe, depends on the size of the gaps. Circlator uses nucmer to match contigs, so as long as nucmer can still find the matches then it will work.

I can't think of anything obvious. Don't really get why it's saying it found `/circlator/build/MUMmer3.23/nucmer` but it thinks the version is < 3.1. Maybe running progcheck with the `--debug` flag...

I would save running time by running circlator after Canu, then polish the output of circlator. Circlator is unlikely to be affected by having input that is not polished.

Yes, it does make sense to align them and take the region of the original assembly, instead of polishing again.

A dot in column 18 means that it is not relevant. If the sample assembles into a contig that has no differences from the reference sequence, then you will see...

Thanks @yjx1217 for the helpful comments! @jhernandez1409 just to expand a bit: you can test if the depenencies are installed by just trying to run them. e.g. just running: bwa...