def main():
option_parser, opts, args = parse_command_line_parameters(**script_info)
if len(opts.taxa_summary_fps) != 2:
option_parser.error("Exactly two taxa summary files are required. You "
"provided %d." % len(opts.taxa_summary_fps))
# Create the output dir if it doesn't already exist.
try:
create_dir(opts.output_dir)
except:
option_parser.error("Could not create or access output directory "
"specified with the -o option.")
sample_id_map = None
if opts.sample_id_map_fp:
sample_id_map = parse_sample_id_map(open(opts.sample_id_map_fp, 'U'))
results = compare_taxa_summaries(
parse_taxa_summary_table(open(opts.taxa_summary_fps[0], 'U')),
parse_taxa_summary_table(open(opts.taxa_summary_fps[1], 'U')),
opts.comparison_mode,
correlation_type=opts.correlation_type,
tail_type=opts.tail_type,
num_permutations=opts.num_permutations,
confidence_level=opts.confidence_level,
perform_detailed_comparisons=opts.perform_detailed_comparisons,
sample_id_map=sample_id_map,
expected_sample_id=opts.expected_sample_id)
# Write out the sorted and filled taxa summaries, basing their
# filenames on the original input filenames. If the filenames are the same,
# append a number to each filename.
same_filenames = False
if basename(opts.taxa_summary_fps[0]) == \
basename(opts.taxa_summary_fps[1]):
same_filenames = True
for orig_ts_fp, filled_ts_lines, file_num in zip(opts.taxa_summary_fps,
results[:2], range(0, 2)):
filename_suffix = '_sorted_and_filled'
if same_filenames:
filename_suffix += '_%d' % file_num
filled_ts_fp = add_filename_suffix(orig_ts_fp, filename_suffix)
filled_ts_f = open(join(opts.output_dir, filled_ts_fp), 'w')
filled_ts_f.write(filled_ts_lines)
filled_ts_f.close()
# Write the overall comparison result.
overall_comp_f = open(join(opts.output_dir, 'overall_comparison.txt'), 'w')
overall_comp_f.write(results[2])
overall_comp_f.close()
# Write the correlation vector containing the pairwise sample comparisons.
if opts.perform_detailed_comparisons:
corr_vec_f = open(join(opts.output_dir, 'detailed_comparisons.txt'),
'w')
corr_vec_f.write(results[3])
corr_vec_f.close()
def main():
option_parser, opts, args = parse_command_line_parameters(**script_info)
if len(opts.taxa_summary_fps) != 2:
option_parser.error("Exactly two taxa summary files are required. You "
"provided %d." % len(opts.taxa_summary_fps))
# Create the output dir if it doesn't already exist.
try:
create_dir(opts.output_dir)
except:
option_parser.error("Could not create or access output directory "
"specified with the -o option.")
sample_id_map = None
if opts.sample_id_map_fp:
sample_id_map = parse_sample_id_map(open(opts.sample_id_map_fp, 'U'))
results = compare_taxa_summaries(
parse_taxa_summary_table(open(opts.taxa_summary_fps[0], 'U')),
parse_taxa_summary_table(open(opts.taxa_summary_fps[1], 'U')),
opts.comparison_mode, correlation_type=opts.correlation_type,
tail_type=opts.tail_type, num_permutations=opts.num_permutations,
confidence_level=opts.confidence_level,
perform_detailed_comparisons=opts.perform_detailed_comparisons,
sample_id_map=sample_id_map,
expected_sample_id=opts.expected_sample_id)
# Write out the sorted and filled taxa summaries, basing their
# filenames on the original input filenames. If the filenames are the same,
# append a number to each filename.
same_filenames = False
if basename(opts.taxa_summary_fps[0]) == \
basename(opts.taxa_summary_fps[1]):
same_filenames = True
for orig_ts_fp, filled_ts_lines, file_num in zip(opts.taxa_summary_fps,
results[:2], range(0, 2)):
filename_suffix = '_sorted_and_filled'
if same_filenames:
filename_suffix += '_%d' % file_num
filled_ts_fp = add_filename_suffix(orig_ts_fp, filename_suffix)
filled_ts_f = open(join(opts.output_dir, filled_ts_fp), 'w')
filled_ts_f.write(filled_ts_lines)
filled_ts_f.close()
# Write the overall comparison result.
overall_comp_f = open(join(opts.output_dir, 'overall_comparison.txt'), 'w')
overall_comp_f.write(results[2])
overall_comp_f.close()
# Write the correlation vector containing the pairwise sample comparisons.
if opts.perform_detailed_comparisons:
corr_vec_f = open(join(opts.output_dir,
'detailed_comparisons.txt'), 'w')
corr_vec_f.write(results[3])
corr_vec_f.close()
def test_valid_get_coefficients_input(self):
"""Functions correctly using standard valid input data"""
exp = ('-0.2336','-0.7924')
run = parse_taxa_summary_table(open(self.L18S_fp, 'U'))
key = parse_taxa_summary_table(open(self.key_fp, 'U'))
obs = get_coefficients(run, key)
self.assertEqual(obs, exp)
def test_valid_get_coefficients_input(self):
"""Functions correctly using standard valid input data"""
exp = ('-0.2336', '-0.7924')
run = parse_taxa_summary_table(open(self.L18S_fp, 'U'))
key = parse_taxa_summary_table(open(self.key_fp, 'U'))
obs = get_coefficients(run, key)
self.assertEqual(obs, exp)
def test_parse_taxa_summary_table(self):
""" parse_taxa_summary_table functions as expected """
actual = parse_taxa_summary_table(self.taxa_summary1.split('\n'))
self.assertEqual(actual[0],self.taxa_summary1_expected[0])
self.assertEqual(actual[1],self.taxa_summary1_expected[1])
self.assertEqual(actual[2],self.taxa_summary1_expected[2])
self.assertEqual(actual,self.taxa_summary1_expected)
def main():
option_parser, opts, args = parse_command_line_parameters(**script_info)
data={}
mapping,headers,comments = get_map(opts, data)
mapping_headers_to_use=opts.mapping_headers_to_use
background_color=opts.background_color
monte_carlo_dist=opts.monte_carlo_dist
ball_scale=opts.ball_scale
arrow_line_color=opts.arrow_line_color
arrow_head_color=opts.arrow_head_color
taxonomy_count_file = opts.input_taxa_file
if taxonomy_count_file:
try:
counts_f = open(taxonomy_count_file, 'U').readlines()
_, taxa_ids, _ = \
parse_taxa_summary_table(counts_f)
except (TypeError, IOError):
raise ValueError, 'Summarized taxa file could not be parsed.'
else:
taxa_ids=None
out = build_prefs_string(mapping_headers_to_use, background_color, \
monte_carlo_dist, headers, taxa_ids, \
ball_scale, arrow_line_color, arrow_head_color)
f = open(opts.output_fp,'w')
f.write(out)
f.close()
def main():
option_parser, opts, args = parse_command_line_parameters(**script_info)
data = {}
mapping, headers, comments = get_map(opts, data)
mapping_headers_to_use = opts.mapping_headers_to_use
background_color = opts.background_color
monte_carlo_dist = opts.monte_carlo_dist
ball_scale = opts.ball_scale
arrow_line_color = opts.arrow_line_color
arrow_head_color = opts.arrow_head_color
taxonomy_count_file = opts.input_taxa_file
if taxonomy_count_file:
try:
counts_f = open(taxonomy_count_file, 'U').readlines()
_, taxa_ids, _ = \
parse_taxa_summary_table(counts_f)
except (TypeError, IOError):
raise ValueError('Summarized taxa file could not be parsed.')
else:
taxa_ids = None
out = build_prefs_string(mapping_headers_to_use, background_color,
monte_carlo_dist, headers, taxa_ids, ball_scale,
arrow_line_color, arrow_head_color)
f = open(opts.output_fp, 'w')
f.write(out)
f.close()
def setUp(self):
"""Set up files/environment that will be used by the tests."""
# The prefix to use for temporary files. This prefix may be added to,
# but all temp dirs and files created by the tests will have this
# prefix at a minimum.
self.prefix = 'generate_taxa_compare_table_tests'
self.start_dir = getcwd()
self.dirs_to_remove = []
self.files_to_remove = []
self.tmp_dir = get_qiime_temp_dir()
if not exists(self.tmp_dir):
makedirs(self.tmp_dir)
# if test creates the temp dir, also remove it
self.dirs_to_remove.append(self.tmp_dir)
# setup temporary root input directory
self.root_dir = mkdtemp(dir=self.tmp_dir,
prefix='%s_root_dir_' %self.prefix)
self.dirs_to_remove.append(self.root_dir)
L18S_dir = '/L18S-1/blast_1.0/'
makedirs(self.root_dir+L18S_dir)
self.L18S_fp = self.root_dir+L18S_dir+'/otu_table_mc2_no_pynast_failures_w_taxa_L5.txt'
with open(self.L18S_fp, 'w') as f:
f.writelines(L18S_L5_blast_one_multiple_assign_output)
self.files_to_remove.append(self.L18S_fp)
# setup temporary key directory
self.key_dir = mkdtemp(dir=self.tmp_dir,
prefix='%s_key_dir_' %self.prefix)
self.dirs_to_remove.append(self.key_dir)
self.key_fp = self.key_dir+'/L18S_key.txt'
with open(self.key_fp, 'w') as f:
f.writelines(L18S_key)
self.files_to_remove.append(self.key_fp)
self.bad_key = self.key_dir+'/L18S_key.txt'
# setup temporary output directory
self.output_dir = mkdtemp(dir=self.tmp_dir,
prefix='%s_output_dir_' %self.prefix)
self.dirs_to_remove.append(self.output_dir)
self.ts1 = parse_taxa_summary_table(ts1.split('\n'))
initiate_timeout(60)
def generate_taxa_compare_table(root, key_directory, levels=None):
"""Finds otu tables in root and compares them against the keys in key_directory.
Walks a file tree starting at root and finds the otu tables output by
multiple_assign_taxonomy.py. Then compares the found otu tables to their corresponding
key in key_directory. Returns a dict containing another dict for every level of output
compared. Output is of the format:
{level: {name of study: {method_and_params: (pearson, spearman)}}}
Parameters:
root: path to root of multiple_assign_taxonomy.py output.
key_directory: path to directory containing known/expected compositions. Each study
should be in its own otu table.
levels: INCOMPLETE. Use other than default will cause unexpected results. The
multiple_assign_taxonomy.py output levels to be analyzed."""
key_fps = get_key_files(key_directory)
results = {}
if not levels:
levels = [2,3,4,5,6]
if len(levels) > 5:
raise WorkflowError('Too many levels.')
for l in levels:
if l < 2 or l > 6:
raise WorkflowError('Level out of range: ' + str(l))
for l in levels:
results[l] = dict()
for(path, dirs, files) in walk(root):
for choice in assignment_method_choices:
#Checks if this dir's name includes a known assignment method (and therefor contains that output)
if choice in path:
study = path.split('/')[-2].rstrip('-123').capitalize()
for f in files:
if 'otu_table_mc2_w_taxa_L' in f and not f.endswith('~'):
name = path.split('/')[-2].capitalize()
level = int(f[-5])
if level not in levels:
#If that level wasn't requested, skip it.
continue
with open(join(path,f),'U') as run_file:
#Open and parse run file
test = run_file.readline()
if('Taxon\t' not in test):
raise WorkflowError('Invalid multiple_assign_taxonomy output file, check for corrupted file: '+path)
run_file.seek(0)
run = parse_taxa_summary_table(run_file)
with open(key_fps[study],'U') as key_file:
#Open and parse key file
test = key_file.readline()
if('Taxon\t' not in test):
raise WorkflowError('Invalid key file in directory: '+path)
key_file.seek(0)
key = parse_taxa_summary_table(key_file)
try:
pearson_coeff, spearman_coeff = get_coefficients(run, key)
except ValueError:
#compare_taxa_summaries couldn't find a match between the 2
#Likely due to mismatch between key and input sample names.
pearson_coeff = 'X'
spearman_coeff = 'X'
try:
results[level][name]
except KeyError:
results[level][name] = dict()
results[level][name][path.split('/')[-1]] = (pearson_coeff, spearman_coeff)
return results
