Method for extracting a simple genus_species string identifier from a BIOM table with taxonomy metadata. If no species, use Genus_spp, if no genus, use Unclassified_ and lowest available taxonomic identifier. This example script will also take a BIOM file from the command line and run the method on every row in the table and print the results.

biom_otu_name.py

import json
import sys

def otu_name(biom_row):
    """
    Determine a simple Genus-species identifier for an OTU, if possible. 
    If OTU is not identified to the species level, name it as Unclassified (familly/genus/etc...)
    """
    tax = biom_row['metadata']['taxonomy']
    for i, lvl in enumerate(tax):
        lvl = lvl.strip()
        if i < len(tax) - 1 and len(tax[i + 1].strip()) == 3:
            if tax[i].strip()[0] == 'g':
                return lvl.split('_')[-1] + '_spp.'
            else:
                return 'Unclassified_' + lvl.split('_')[-1]
        elif i == len(tax) - 1:
            name = lvl.split('_')[-1]
            if lvl[0] == 's':
                name = tax[i-1].split('_')[-1] + '_' + name
            return name
            
if __name__ == '__main__':
    with open(sys.argv[1], 'rU') as biomF:
        biom = json.loads(biomF.readline())
    
    for row in biom['rows']:
        print otu_name(row)