def musicc_normalization(profile, in_file, output_dir, musicc_inter=True, input_format='tab', output_format='tab',
musicc_intra='use_generic', compute_scores=False, verbose=False):
"""Perform a MUSiCC normalization on this profile's abundance data. Write normalized
data to file and update profile's abundance data to be normalized.
Requires:
MUSiCC module can be imported and abundance features are KOs.
Args:
profile (metagenomic_profile): profile containing the abundance data
in_file (str): path to original abundance data file
output_dir (str): output directory where normalized data will be saved
Effects:
profile's abundance data is now normalized.
Note:
For more details see "MUSiCC: A marker genes based framework for
metagenomic normalization and accurate profiling of gene abundances in the
microbiome." Ohad Manor and Elhanan Borenstein. Genome Biology.
"""
from musicc.core import correct_and_normalize
musicc_args = {'musicc_inter':musicc_inter, 'input_format':input_format, 'output_format':output_format,
'musicc_intra':musicc_intra, 'compute_scores':compute_scores, 'verbose':verbose}
musicc_args['input_file'] = in_file
musicc_args['output_file'] = output_dir + "//" + "musicc_normalized_abundance.tab"
correct_and_normalize(musicc_args)
profile.set_abundance_data(pd.DataFrame.from_csv(musicc_args['output_file'], sep='\t'))
def test_is_output_correct_for_normalization_correction_learn_model(self):
"""Does MUSiCC produce the correct output for normalization and correction of the example case?"""
# define the arguments needed by MUSiCC
musicc_args = {
'input_file': MUSiCCTestCase.path_to_data +
'/examples/simulated_ko_relative_abundance.tab',
'output_file': MUSiCCTestCase.path_to_data + '/examples/test3.tab',
'input_format': 'tab',
'output_format': 'tab',
'musicc_inter': True,
'musicc_intra': 'learn_model',
'compute_scores': True,
'verbose': False
}
# run the MUSiCC correction
correct_and_normalize(musicc_args)
# assert that the result is equal to the example (up to small difference due to de novo learning)
example = pd.read_table(
MUSiCCTestCase.path_to_data +
'/examples/simulated_ko_MUSiCC_Normalized_Corrected_learn_model.tab',
index_col=0)
output = pd.read_table(MUSiCCTestCase.path_to_data +
'/examples/test3.tab',
index_col=0)
example_vals = example.values
output_vals = output.values
self.assertTrue(example_vals.shape[0] == output_vals.shape[0])
self.assertTrue(example_vals.shape[1] == output_vals.shape[1])
for i in range(example_vals.shape[0]):
for j in range(example_vals.shape[1]):
self.assertTrue(
abs(example_vals[i, j] - output_vals[i, j]) < 1)
os.remove(MUSiCCTestCase.path_to_data + '/examples/test3.tab')
def musicc_normalization(profile,
in_file,
output_dir,
musicc_inter=True,
input_format='tab',
output_format='tab',
musicc_intra='use_generic',
compute_scores=False,
verbose=False):
"""Perform a MUSiCC normalization on this profile's abundance data. Write normalized
data to file and update profile's abundance data to be normalized.
Requires:
MUSiCC module can be imported and abundance features are KOs.
Args:
profile (metagenomic_profile): profile containing the abundance data
in_file (str): path to original abundance data file
output_dir (str): output directory where normalized data will be saved
Effects:
profile's abundance data is now normalized.
Note:
For more details see "MUSiCC: A marker genes based framework for
metagenomic normalization and accurate profiling of gene abundances in the
microbiome." Ohad Manor and Elhanan Borenstein. Genome Biology.
"""
from musicc.core import correct_and_normalize
musicc_args = {
'musicc_inter': musicc_inter,
'input_format': input_format,
'output_format': output_format,
'musicc_intra': musicc_intra,
'compute_scores': compute_scores,
'verbose': verbose
}
musicc_args['input_file'] = in_file
musicc_args[
'output_file'] = output_dir + "//" + "musicc_normalized_abundance.tab"
correct_and_normalize(musicc_args)
profile.set_abundance_data(
pd.DataFrame.from_csv(musicc_args['output_file'], sep='\t'))
def test_is_output_correct_for_normalization_correction_learn_model(self):
"""Does MUSiCC produce the correct output for normalization and correction of the example case?"""
# define the arguments needed by MUSiCC
musicc_args = {'input_file': MUSiCCTestCase.path_to_data + '/examples/simulated_ko_relative_abundance.tab',
'output_file': MUSiCCTestCase.path_to_data + '/examples/test3.tab',
'input_format': 'tab', 'output_format': 'tab', 'musicc_inter': True,
'musicc_intra': 'learn_model', 'compute_scores': True, 'verbose': False}
# run the MUSiCC correction
correct_and_normalize(musicc_args)
# assert that the result is equal to the example (up to small difference due to de novo learning)
example = pd.read_table(MUSiCCTestCase.path_to_data + '/examples/simulated_ko_MUSiCC_Normalized_Corrected_learn_model.tab', index_col=0)
output = pd.read_table(MUSiCCTestCase.path_to_data + '/examples/test3.tab', index_col=0)
example_vals = example.values
output_vals = output.values
self.assertTrue(example_vals.shape[0] == output_vals.shape[0])
self.assertTrue(example_vals.shape[1] == output_vals.shape[1])
for i in range(example_vals.shape[0]):
for j in range(example_vals.shape[1]):
self.assertTrue(abs(example_vals[i, j] - output_vals[i, j]) < 1)
os.remove(MUSiCCTestCase.path_to_data + '/examples/test3.tab')
default='tab')
parser.add_argument('-n',
'--normalize',
dest='musicc_inter',
help='Apply MUSiCC normalization (default: false)',
action='store_true')
parser.add_argument(
'-c',
'--correct',
dest='musicc_intra',
choices=['use_generic', 'learn_model'],
help='Correct abundance per-sample using MUSiCC (default: false)',
default='None')
parser.add_argument(
'-perf',
'--performance',
dest='compute_scores',
help=
'Calculate model performance on various gene sets (may add to running time) (default: false)',
action='store_true')
parser.add_argument('-v',
'--verbose',
dest='verbose',
help='Increase verbosity of module (default: false)',
action='store_true')
given_args = parser.parse_args()
# run normalization and correction
correct_and_normalize(vars(given_args))
#!/usr/bin/env python
import argparse
# for testing the new module, remove later!!!!!!
import sys
sys.path.append('/net/gs/vol1/home/ohadm/MUSiCC/PyCode/MUSiCC')
from musicc.core import correct_and_normalize
if __name__ == "__main__":
# get options from user
parser = argparse.ArgumentParser(description='MUSiCC: Metagenomic Universal Single-Copy Correction')
parser.add_argument('input_file', help='Input abundance file to correct')
parser.add_argument('-o', '--out', dest='output_file', help='Output destination for corrected abundance (default: MUSiCC.tab)', default='MUSiCC.tab')
parser.add_argument('-if', '--input_format', dest='input_format', choices=['tab', 'csv', 'biom'], help='Option indicating the format of the input file (default: tab)', default='tab')
parser.add_argument('-of', '--output_format', dest='output_format', choices=['tab', 'csv', 'biom'], help='Option indicating the format of the output file (default: tab)', default='tab')
parser.add_argument('-n', '--normalize', dest='musicc_inter', help='Apply MUSiCC normalization (default: false)', action='store_true')
parser.add_argument('-c', '--correct', dest='musicc_intra', choices=['use_generic', 'learn_model'], help='Correct abundance per-sample using MUSiCC (default: false)', default='None')
parser.add_argument('-perf', '--performance', dest='compute_scores', help='Calculate model performance on various gene sets (may add to running time) (default: false)', action='store_true')
parser.add_argument('-v', '--verbose', dest='verbose', help='Increase verbosity of module (default: false)', action='store_true')
given_args = parser.parse_args()
# run normalization and correction
correct_and_normalize(vars(given_args))
