def test_parse_metadata_state_descriptions(self):
"""parse_metadata_state_descriptions should return correct states from string."""
s = ''
self.assertEqual(parse_metadata_state_descriptions(s), {})
s = 'Study:Twin,Hand,Dog;BodySite:Palm,Stool'
self.assertEqual(parse_metadata_state_descriptions(s), {'Study':set(['Twin','Hand','Dog']),
'BodySite':set(['Palm','Stool'])})
def get_sam_ids(map_data, map_header, colorby, cat, primary_state, secondary_state):
""" returns all sample ids matching the state strings and colorby:cat
colorby: eg: 'Country', or pass None to not filter only colorby:cat samples
cat: e.g.: 'USA'
primary_state: e.g.: 'AgeCategory:Child'
secondary state can be None, or like primary state
returns uniquified lists in randomized order
"""
if colorby == None:
sample_ids = [sam[0] for sam in map_data]
else:
sample_ids = get_sample_ids(map_data, map_header, {colorby: [cat]})
# primary key is the category label, e.g. AgeCategory
# value is the val for that category, e.g. Adult
# go through age1/age2
primary_states = parse_metadata_state_descriptions(primary_state)
if colorby != None:
primary_states[colorby] = [cat]
state1_samids = get_sample_ids(map_data, map_header, primary_states)
if secondary_state == None:
state2_samids = set(sample_ids).difference(set(state1_samids))
else:
secondary_states = parse_metadata_state_descriptions(secondary_state)
if colorby != None:
secondary_states[colorby] = [cat]
state2_samids = get_sample_ids(map_data, map_header, secondary_states)
return list(set(state1_samids)), list(set(state2_samids))
def get_sam_ids(map_data, map_header, colorby, cat, primary_state,
secondary_state):
""" returns all sample ids matching the state strings and colorby:cat
colorby: eg: 'Country', or pass None to not filter only colorby:cat samples
cat: e.g.: 'USA'
primary_state: e.g.: 'AgeCategory:Child'
secondary state can be None, or like primary state
returns uniquified lists in randomized order
"""
if colorby is None:
sample_ids = [sam[0] for sam in map_data]
else:
sample_ids = get_sample_ids(map_data, map_header, {colorby: [cat]})
# primary key is the category label, e.g. AgeCategory
# value is the val for that category, e.g. Adult
# go through age1/age2
primary_states = parse_metadata_state_descriptions(primary_state)
if colorby is not None:
primary_states[colorby] = [cat]
state1_samids = get_sample_ids(map_data, map_header, primary_states)
if secondary_state is None:
state2_samids = set(sample_ids).difference(set(state1_samids))
else:
secondary_states =\
parse_metadata_state_descriptions(secondary_state)
if colorby is not None:
secondary_states[colorby] = [cat]
state2_samids = get_sample_ids(map_data, map_header, secondary_states)
return list(set(state1_samids)), list(set(state2_samids))
def test_parse_metadata_state_descriptions(self):
"""parse_metadata_state_descriptions should return correct states from string."""
s = ''
self.assertEqual(parse_metadata_state_descriptions(s), {})
s = 'Study:Twin,Hand,Dog;BodySite:Palm,Stool'
self.assertEqual(
parse_metadata_state_descriptions(s), {
'Study': set(['Twin', 'Hand', 'Dog']),
'BodySite': set(['Palm', 'Stool'])
})
def test_get_sample_ids(self):
"""get_sample_ids should return sample ids matching criteria."""
self.assertEqual(get_sample_ids(self.map_data, self.map_headers,\
parse_metadata_state_descriptions('Study:Twin')), [])
self.assertEqual(get_sample_ids(self.map_data, self.map_headers,\
parse_metadata_state_descriptions('Study:Dog')), ['a','b'])
self.assertEqual(get_sample_ids(self.map_data, self.map_headers,\
parse_metadata_state_descriptions('Study:*,!Dog')), ['c','d','e'])
self.assertEqual(get_sample_ids(self.map_data, self.map_headers,\
parse_metadata_state_descriptions('Study:*,!Dog;BodySite:Stool')), ['e'])
self.assertEqual(get_sample_ids(self.map_data, self.map_headers,\
parse_metadata_state_descriptions('BodySite:Stool')), ['a','b','e'])
def sample_ids_from_metadata_description(mapping_f,valid_states_str):
""" Given a description of metadata, return the corresponding sample ids
"""
map_data, map_header, map_comments = parse_mapping_file(mapping_f)
valid_states = parse_metadata_state_descriptions(valid_states_str)
sample_ids = get_sample_ids(map_data, map_header, valid_states)
return sample_ids
def filter_otus_and_map(map_infile, otu_infile, map_outfile, otu_outfile,
valid_states_str, num_seqs_per_otu):
"""Filters OTU and map files according to specified criteria."""
map_data, map_header, map_comments = parse_mapping_file(map_infile)
map_infile.close()
valid_states = parse_metadata_state_descriptions(valid_states_str)
sample_ids = get_sample_ids(map_data, map_header, valid_states)
# write out the filtered mapping file
out_headers, out_data = filter_map(map_data, map_header, sample_ids)
header_line = '#' + '\t'.join(out_headers)
map_outfile.write('\n'.join([header_line] + map('\t'.join, out_data)))
if not isinstance(map_outfile, StringIO):
map_outfile.close()
# write out the filtered OTU file
for line in otu_infile:
if line.startswith('#OTU ID'):
fields = map(strip, line.split('\t'))
cols = find_good_cols(line, sample_ids)
filter_line(line, cols, min_count=None, outfile=otu_outfile)
elif line.startswith('#'):
otu_outfile.write(line)
else:
filter_line(line, cols, min_count=num_seqs_per_otu,
outfile=otu_outfile)
if not isinstance(otu_outfile, StringIO):
otu_outfile.close()
def sample_ids_from_metadata_description(mapping_f, valid_states_str):
""" Given a description of metadata, return the corresponding sample ids
"""
map_data, map_header, map_comments = parse_mapping_file(mapping_f)
valid_states = parse_metadata_state_descriptions(valid_states_str)
sample_ids = get_sample_ids(map_data, map_header, valid_states)
return sample_ids
def filter_otus_and_map(map_infile, otu_infile, map_outfile, otu_outfile,
valid_states_str, num_seqs_per_otu):
"""Filters OTU and map files according to specified criteria."""
map_data, map_header, map_comments = parse_mapping_file(map_infile)
map_infile.close()
valid_states = parse_metadata_state_descriptions(valid_states_str)
sample_ids = get_sample_ids(map_data, map_header, valid_states)
# write out the filtered mapping file
out_headers, out_data = filter_map(map_data, map_header, sample_ids)
header_line = '#' + '\t'.join(out_headers)
map_outfile.write('\n'.join([header_line] + map('\t'.join, out_data)))
if not isinstance(map_outfile, StringIO):
map_outfile.close()
# write out the filtered OTU file
for line in otu_infile:
if line.startswith('#OTU ID'):
fields = map(strip, line.split('\t'))
cols = find_good_cols(line, sample_ids)
filter_line(line, cols, min_count=None, outfile=otu_outfile)
elif line.startswith('#'):
otu_outfile.write(line)
else:
filter_line(line, cols, min_count=num_seqs_per_otu,
outfile=otu_outfile)
if not isinstance(otu_outfile, StringIO):
otu_outfile.close()
def sample_ids_from_metadata_description(mapping_f, valid_states_str):
""" Given a description of metadata, return the corresponding sample ids
"""
map_data, map_header, map_comments = parse_mapping_file(mapping_f)
valid_states = parse_metadata_state_descriptions(valid_states_str)
sample_ids = get_sample_ids(map_data, map_header, valid_states)
if len(sample_ids) < 1:
raise ValueError("All samples have been filtered out for the criteria" " described in the valid states")
return sample_ids
def sample_ids_from_metadata_description(mapping_f, valid_states_str):
""" Given a description of metadata, return the corresponding sample ids
"""
map_data, map_header, map_comments = parse_mapping_file(mapping_f)
valid_states = parse_metadata_state_descriptions(valid_states_str)
sample_ids = get_sample_ids(map_data, map_header, valid_states)
if len(sample_ids) < 1:
raise ValueError,"All samples have been filtered out for the criteria"+\
" described in the valid states"
return sample_ids
