How to get Greengenes taxonomies

Find the database

You'll need to find the file that links the Greengenes IDs with the taxonomies. For example, gg_13_5/gg_13_5_taxonomy.txt.

Searching the plain text file is slow, so you can create a fast look-up hash table using the Smile Train script util/setup_tools/pickle_taxonomy.py. I recommend placing the pickled file in, e.g., gg_13_5/pickles/gg_13_5_taxonomy.pkl.

Getting the taxonomies

You can use the Smile Train script util/get_taxonomies.py. For example, to get the Greengenes taxonomies for an OTU table otu.txt that has Greengenes IDs and put them in tax.txt, you would call

/path/to/SmileTrain/util/get_taxonomies.py otu.txt /path/to/gg_13_5/pickles/gg_13_5_taxonomy.pkl > tax.txt

If you are using a pickled dictionary, this is fast, so you can just run it on the head node of a compute cluster. If you have not pickled the taxonomy list, it would be impolite to run on the head node.