parameter to produce one taxonomy ID per sequence about centrifuge HOT 4 CLOSED

One option is to use parameter "-k 1".
As a result, more reads would be assigned to higher-level in the taxonomy tree comparing using default -k value. But some reads might be assigned to totally different branch in the taxonomy tree. We will make the result more consistent when changing "-k" value in future version.
Is this what you need? Thanks.

from centrifuge.

Thanks a lot, that's exactly what I need.

from centrifuge.

Hi,

I have met another problem while running centrifuge with "-k 1", It seems that not all reads have a corresponding taxonomic id in result file, I think centrifuge doesn't output those reads with bad classification result, however I need those reads appear in result file, so is there a way to output all reads into classification file even though some of them may not have a taxonomic id?

Thanks

from centrifuge.

This feature is not implemented in current version.
You can use "--reorder" option to force the output to have the same order as the input fasta/q file. Then you can figure out the unassigned reads and modify the report accordingly.

from centrifuge.