reorient/demux mixed-orientation reads about q2-cutadapt HOT 8 CLOSED

It looks like we've decided on something resembling @nbokulich's possible solution 1 where the reorientation and the demuxing are done together in the same step. This decision is based on the behavior described in the link to the cutadapt documentation @thermokarst linked above. We've come up with a couple possible ways of implementing this.

1. Add a --reorient flag to the existing demux_paired command that, when provided, would demux the user's reads as normal, then reorient them and demux again
2. Create a new demux_paired_reorient that would implement the behavior described above in its own entirely separate method
   Making this the default behavior feels excessive as it could significantly slow down the demuxing process. @thermokarst and I also discussed the possible issue of how to let the user know this is something they might want to do i.e. how do they know if their data is formatted in a manner that would require this? Any thoughts or comments on any of the above are appreciated.

from q2-cutadapt.

I think I am leaning more towards a param on the existing demux_paired method --- maybe mixed_orientation: bool = False?

from q2-cutadapt.

forum x-ref

from q2-cutadapt.

forum xref

from q2-cutadapt.

forum xref

from q2-cutadapt.

A cutadapt protocol is outlined here:

https://cutadapt.readthedocs.io/en/stable/guide.html?highlight=orientation#demultiplexing-paired-end-reads-in-mixed-orientation

from q2-cutadapt.

@thermokarst I think I prefer that idea too

from q2-cutadapt.

This just came up on the forum again https://forum.qiime2.org/t/cutadapt-demultiplexing-with-mixed-primers/15260/2

from q2-cutadapt.