LANG: add new pathogenome distro

[lizgehret]

opening a LANG PR since there are requisite changes on q2-amr and genome sampler that need to be tested concurrently.

• LANG: recipe updates for pathogenome distro bokulich-lab/q2-rgi#103

NOTE: genome-sampler changes are on hold until i have write access (so a branch can be created on the canonical remote)

[lizgehret]

Okay so the main conflict right now is the rgi pin in q2-amr - there aren't any linux/osx builds of this on bioconda that surpass python 3.6 (the last non-noarch builds were uploaded almost five years ago).

[lizgehret]

Alright so latest updates are that rgi pins a very old version of pysam (0.16.0.1) that's not compatible with python 3. I've opened an issue on their repo to request for the pin to be relaxed as a min pin, so we'll see what happens. Not sure if we have another way around this in the meantime unless it's possible for q2-amr to remove the rgi pin.
(EDIT: doesn't seem likely after looking over q2-amr's repo - it pretty heavily depends on rgi)

[lizgehret]

Necessary updates have been made - q2-amr has been split into q2-amrfinderplus & q2-rgi, which should resolve conflicts. i've updated the recipe to account for this plugin change, and remove stats (as it is no longer a dependency for vizard).

[lizgehret]

package perl-module-runtime-0.016-pl5321hdfd78af_2 is excluded by strict repo priority

metapackage failed to solve due to strict channel (?) priority not allowing a dependency to be included. i couldn't find any info on strict repo priority, so i'm guessing this is the channel priority. re-running with the defaults channel included to see what happens.

[lizgehret]

• q2-amrfinderplus had a stale name in the recipe file (q2-amr vs. amrfinderplus) - paired PR to address this here.
• There's an apparent incompatibility with q2-viromics' python 3.10 ubuntu build (osx build is fine) that's causing the test failure (@Oddant1 was able to repro locally). I'm going to work on addressing that today.
• RESCRIPt test is still failing, but is difficult to test locally (tests pass locally & when RESCRIPt is built in isolation). Also on my to-do list for today.
• feature-classifier test is failing on osx, and it seems like the culprit may be a different version of vsearch. added a version pin in this PR to address the issue.

[lizgehret]

viromics PR to fix failure: bokulich-lab/q2-viromics#19

[lizgehret]

RESCRIPt PR to fix test failure: bokulich-lab/RESCRIPt#206

[lizgehret]

🥳🥳🥳