Use this script to add taxonomy strings to feature importance scores from OTU level data (see https://gist.github.com/walterst/2222618976a66b3fc8dd)

get_matching_taxa.py

#!/usr/bin/env python

# USAGE: python get_matching_taxa.py tab_sep_OTU_table feature_importance_file > feature_imp_with_taxa.txt

from sys import argv

otu_table = open(argv[1], "U")

otu_taxa = {}

for line in otu_table:
    if line.startswith("#"):
        continue
    if len(line.strip()) == 0:
        continue
    curr_line = line.split("\t")
    otu = curr_line[0]
    taxa = curr_line[-1].strip()
    
    otu_taxa[otu] = taxa
    
importance_table = open(argv[2], "U")

for line in importance_table:
    curr_line = line.strip().replace('"', '').split('\t')
    if line.startswith('"CCD"'):
        print '%s\tTaxonomy' % line.strip()
        continue
    try:
        print "%s\t%s" % ("\t".join(curr_line), otu_taxa[curr_line[0]].replace(" ", ""))
    except KeyError:
        print "%s\tNA" % "\t".join(curr_line)