search

nlapier2 / Metalign

Metalign: efficient alignment-based metagenomic profiling via containment min hash
MIT License
32 stars 7 forks source link

setup_data.sh not in bioconda release #35

Open dkoslicki opened 4 years ago

dkoslicki commented 4 years ago

Appears only python scripts were added. Will need to modify setup.py so setup_data.sh is accessible on Bioconda install

dkoslicki commented 4 years ago

Note: removing this if statement in setup.py does make bioconda install the setup_data.sh script. However, @nlapier2 as noted in #34, since setup_data.sh could be run from anywhere, if after bioconda install, I run setup_data.sh, it ends up getting dumped wherever I happened to be when I ran the command.

I suggest the following:

  1. Make setup_data.sh require the user to put a directory <dir> where they want the data downloaded. I suggest using the same flag --db_dir so the user sees clearly how it ties in with metalign.py.
  2. Put spec_db_info.txt on box as well (as this is tied to the default training data anyways) so it (along with all training data) gets put in <dir>
  3. In metalign.py, leave --db_dir but remove --dbinfo_in (as from 2 above, it will already be inside whatever directory the user specified in 1).