def test_growth_rates(self, data_folder, test_folder):
model_files = ['BT.sbml', 'FP.sbml']
source_models = mminte.get_all_pairs(
[join(data_folder, x) for x in model_files])
pair_models = mminte.create_interaction_models(source_models,
test_folder)
assert len(pair_models) == 1
assert pair_models[0] == '{0}/BTxFP.json'.format(test_folder) or \
pair_models[0] == '{0}/FPxBT.json'.format(test_folder)
western = json.load(open(join(data_folder, 'western.json')))
growth_rates = mminte.calculate_growth_rates(pair_models, western)
assert growth_rates.at[0, 'A_ID'] == 'BT'
assert growth_rates.at[0, 'B_ID'] == 'FP'
assert growth_rates.at[0, 'TYPE'] == 'Parasitism'
assert growth_rates.at[0, 'TOGETHER'] == pytest.approx(0.49507501)
assert growth_rates.at[0, 'A_TOGETHER'] == pytest.approx(0.27746256)
assert growth_rates.at[0, 'B_TOGETHER'] == pytest.approx(0.21761245)
assert growth_rates.at[0, 'A_ALONE'] == pytest.approx(0.44073842)
assert growth_rates.at[0, 'B_ALONE'] == pytest.approx(0.16933796)
assert growth_rates.at[0, 'A_CHANGE'] == pytest.approx(-0.37045977)
assert growth_rates.at[0, 'B_CHANGE'] == pytest.approx(0.28507777)
for model in pair_models:
unlink(model)
def prepare(single_file_names,
pair_model_folder,
optimize=False,
n_processes=None):
""" Prepare for simulation by creating two species community models.
Parameters
----------
single_file_names : list of str
Paths to single species model files of community members
pair_model_folder : str
Path to folder for storing two species community models
optimize : bool
When True, confirm single species models optimize successfully before creating community models
n_processes: int, optional
Number of processes in job pool
Returns
-------
list of str
List of paths to two species community models
"""
# If requested, confirm all of the single species input models produce growth as provided.
if optimize:
LOGGER.info('Started checking %d single models for growth',
len(single_file_names))
if n_processes is None:
n_processes = min(cpu_count(), 4)
pool = Pool(n_processes)
result_list = [
pool.apply_async(check_for_growth, (file_name, ))
for file_name in single_file_names
]
for result in result_list:
summary = result.get()
if not summary['grows']:
warn(summary['message'])
LOGGER.info('Finished checking single models for growth')
# Create folder for output pair models if needed.
if not exists(pair_model_folder):
makedirs(pair_model_folder)
# Create all of the pair models and store in specified folder.
LOGGER.info('Started creating two species community models')
pair_file_names = create_interaction_models(
get_all_pairs(single_file_names),
pair_model_folder,
n_processes=n_processes)
LOGGER.info('Finished creating two species community models')
return pair_file_names
def getHTML(self, params):
""" Run Widget All and generate HTML output for Results tab. """
# Validate input parameters.
cherrypy.log('Widget A input parameters: {0}'.format(params))
if not exists(params['correlation_file']):
cherrypy.log('Widget A: correlation file "{0}" was not found'.format(params['correlation_file']))
return 'Sorry, correlation file "{0}" was not found. Make sure the path to the file is correct.' \
.format(params['correlation_file'])
if not exists(params['representative_otu_file']):
cherrypy.log('Widget A: representative OTU file "{0}" was not found'
.format(params['representative_otu_file']))
return 'Sorry, representative OTU file "{0}" was not found. Make sure the path to the file is correct.' \
.format(params['representative_otu_file'])
if not exists(params['diet_file']):
cherrypy.log('Widget A: diet file "{0}" was not found'.format(params['diet_file']))
return 'Sorry, diet file "{0}" was not found. Make sure the path to the file is correct.' \
.format(params['diet_file'])
if not exists(params['analysis_folder']):
try:
makedirs(params['analysis_folder'])
except Exception as e:
cherrypy.log('Widget A: Error creating folder "{0}" for analysis files: {1}'
.format(params['analysis_folder'], e))
return 'Sorry, something went wrong creating the folder "{0}" for the analysis files. Make sure ' \
'the path to the file is correct.<br><br>Exception: {1}'.format(params['analysis_folder'], e)
# Widget 1 - Get the unique OTU sequences.
try:
cherrypy.log('Widget A1: Started getting unique OTU sequences')
unique_otus_file = join(params['analysis_folder'], 'unique_otus.fasta')
get_unique_otu_sequences(read_correlation_file(params['correlation_file']),
params['representative_otu_file'],
unique_otus_file)
cherrypy.log("Widget A1: Finished getting unique OTU sequences")
except Exception as e:
cherrypy.log('Widget A1: Error getting unique OTU sequences: {0}'.format(e))
return "Sorry, something went wrong. Make sure the paths to your files are correct.<br><br>" \
"Exception: {0}.".format(e)
# Widget 2 - Run blast search to find matching bacterial species.
try:
cherrypy.log('Widget A2: Started blast search for matching bacterial species')
blast_output_file = join(params['analysis_folder'], 'blast.txt')
genome_ids, similarity = search(unique_otus_file, blast_output_file)
with open(join(params['analysis_folder'], 'genome_ids.txt'), 'w') as handle:
handle.write('\n'.join(genome_ids)+'\n')
write_similarity_file(similarity, join(params['analysis_folder'], 'similarity.csv'))
cherrypy.log("Widget A2: Finished blast search")
except Exception as e:
cherrypy.log("Widget A2: Error running blast search: {0}".format(e))
return "Sorry, something went wrong. Make sure the paths to your files are correct and " \
"that the correct version of blast is installed.<br><br>Exception: {0}".format(e)
# Widget 3 - Create single species models using ModelSEED.
model_folder = join(params['analysis_folder'], 'single_models')
if not exists(model_folder):
try:
makedirs(model_folder)
except Exception as e:
cherrypy.log('Widget A3: Error creating folder "{0}" for model files: {1}'
.format(params['model_folder'], e))
return 'Sorry something went wrong creating the folder "{0}" for the model files. Make sure ' \
'the path to the folder is correct.<br><br>Exception: {1}'.format(params['model_folder'], e)
try:
cherrypy.log('Widget A3: Started creating models for {0} genomes'.format(len(genome_ids)))
single_filenames = create_species_models(genome_ids, model_folder)
output_filename = join(params['analysis_folder'], 'single_model_filenames.txt')
with open(output_filename, 'w') as handle:
handle.write('\n'.join(single_filenames)+'\n')
cherrypy.log('Widget A3: Finished creating and downloading {0} models'.format(len(single_filenames)))
except Exception as e:
cherrypy.log('Widget A3: Error creating models: {0}'.format(e))
return "Sorry, something went wrong creating metabolic models using ModelSEED.<br><br>" \
"Exception: {0}".format(e)
# Widget 4 - Create two species community models.
pair_model_folder = join(params['analysis_folder'], 'pair_models')
if not exists(pair_model_folder):
try:
makedirs(pair_model_folder)
except Exception as e:
cherrypy.log('Widget A4: We were unable to create folder "{0}" for community model files'
.format(pair_model_folder))
return 'Sorry, something went wrong creating the folder "{0}" for the community model files. ' \
'Make sure the path to the folder is correct.<br><br>Exception: {1}' \
.format(pair_model_folder, e)
try:
pairs = get_all_pairs(single_filenames)
cherrypy.log('Widget A4: Started creating community models from {0} pairs of single species models'
.format(len(pairs)))
pair_filenames = create_interaction_models(pairs, output_folder=pair_model_folder)
output_filename = join(params['analysis_folder'], 'pair_model_filenames.txt')
with open(output_filename, 'w') as handle:
handle.write('\n'.join(pair_filenames)+'\n')
cherrypy.log("Widget A4: Finished creating {0} community models".format(len(pair_filenames)))
except Exception as e:
cherrypy.log("Widget A4: Error creating community models: {0}".format(e))
return "Sorry, something went wrong. Make sure the path to your file is correct and " \
"that the Python package cobrapy is loaded into your system.<br><br>Exception: {0}".format(e)
# Widget 5 - Calculate growth rates for the two species models.
try:
cherrypy.log("Widget A5: Started calculating growth rates for {0} pair models"
.format(len(pair_filenames)))
medium = read_diet_file(params['diet_file'])
growth_rates = calculate_growth_rates(pair_filenames, medium)
write_growth_rates_file(growth_rates, join(params['analysis_folder'], 'growth_rates.csv'))
cherrypy.log("Widget A5: Finished calculating the growth rates")
except Exception as e:
cherrypy.log("Widget A5: Error calculating growth rates: {0}".format(e))
return "Sorry, something went wrong. Make sure the path to your file is correct.<br><br>" \
"Exception: {0}".format(e)
# Widget 6 - Generate data for plot of interaction network.
try:
cherrypy.log('Widget A6: Started generating data for plot of interaction network')
correlation = read_correlation_file(params['correlation_file'])
make_d3_source(growth_rates, join(params['analysis_folder'], 'data4plot.json'), similarity, correlation)
make_d3_source(growth_rates, self.getRoot().data4plot_filename(), similarity, correlation)
cherrypy.log('Widget A6: Finished generating data for plot of interaction network')
except Exception as e:
cherrypy.log('Widget A6: Error generating data for plot of network: {0}'.format(e))
return 'Sorry, something went wrong. Make sure the locations of your files are correct.<br><br>' \
'Exception: {0}'.format(e)
# Generate the output for the Results tab.
text = ["The plot with the network of interactions between your organisms is shown below or "
"on a new tab.<br><br>The shading of the nodes indicates how close the sequence of the "
"OTU is to the sequence of the genome. The darker the node, the higher the similarity.<br><br>"
"The length and thickness of the links reflect the association values on the initial "
"file you provided. The shorter and thicker the link, the higher the association value.<br><br>"
"The colors of the links reflect the kind of interaction. The red, green and grey "
"represent negative, positive and no interaction, respectively.<br><br>"
"<a href='http://d3js.org/'>D3 is awesome</a>! If you mouse over the nodes, you get "
"the ID of the OTU, and if you click a node and drag it, the network will follow it."]
if params['browser_tab'] == 'Current':
text.append(self.getRoot().widget6_out())
else:
webbrowser.open('http://localhost:8080/widget6_out', new=1)
return text
def getHTML(self, params):
""" Run Widget 4 and generate HTML output for Results tab. """
# Validate input parameters.
cherrypy.log('Widget 4 input parameters: {0}'.format(params))
if not exists(params['analysis_folder']):
try:
makedirs(params['analysis_folder'])
except Exception as e:
cherrypy.log('Widget 4: Error creating folder "{0}" for analysis files: {1}'
.format(params['analysis_folder'], e))
return 'Sorry, something went wrong creating the folder "{0}" for the analysis files. Make sure ' \
'the path to the file is correct.<br><br>Exception: {1}'.format(params['analysis_folder'], e)
community_folder = join(params['analysis_folder'], params['community_folder'])
if not exists(community_folder):
try:
makedirs(community_folder)
except Exception as e:
cherrypy.log('Widget 4: We were unable to create folder "{0}" for community model files'
.format(community_folder))
return 'Sorry, something went wrong creating the folder "{0}" for the community model files. ' \
'Make sure the path to the folder is correct.<br><br>Exception: {1}' \
.format(community_folder, e)
if params['pair_input_type'] == 'all':
model_list_file = join(params['analysis_folder'], params['single_models_file'])
if not exists(model_list_file):
cherrypy.log('Widget 4: Model list file "{0}" was not found'.format(model_list_file))
return 'Sorry, model list file "{0}" was not found. Make sure the path to the file is correct.' \
.format(model_list_file)
with open(model_list_file, 'r') as handle:
source_models = [line.strip() for line in handle]
pairs = get_all_pairs(source_models)
cherrypy.log('Widget 4: Generated {0} pairs from {1} source models'.format(len(pairs), len(source_models)))
elif params['pair_input_type'] == 'specific':
pair_list_file = join(params['analysis_folder'], params['pair_list_file'])
if not exists(pair_list_file):
cherrypy.log('Widget 4: Pair list file "{0}" was not found'.format(pair_list_file))
return 'Sorry, pair list file "{0}" was not found. Make sure the path to the file is correct.' \
.format(pair_list_file)
with open(pair_list_file, 'r') as handle:
pair_list = [line.strip().split('\t') for line in handle]
for index in range(len(pair_list)):
fields = pair_list[index]
if len(fields) != 2:
return 'Line {0} in "{1}" file must have two columns separated by tab' \
.format(index, params['pair_list_file'])
pairs = [(fields[0], fields[1]) for fields in pair_list]
cherrypy.log('Widget 4: Found {0} pairs in pair list file'.format(len(pairs)))
else:
similarity_file = join(params['analysis_folder'], params['similarity_file'])
if not exists(similarity_file):
cherrypy.log('Widget 4: Similarity file "{0}" was not found'.format(similarity_file))
return 'Sorry, similarity file "{0}" was not found. Make sure the path to the file is correct.' \
.format(similarity_file)
if not exists(params['correlation_file']):
cherrypy.log('Widget 6: correlation file "{0}" was not found'.format(params['correlation_file']))
return 'Sorry, correlation file "{0}" was not found. Make sure the path to the file is correct.' \
.format(params['correlation_file'])
similarity = read_similarity_file(similarity_file)
correlation = read_correlation_file(params['correlation_file'])
pairs = list()
for corr in correlation:
one = similarity.loc[similarity['OTU_ID'] == corr[0]].iloc[0]['GENOME_ID']
two = similarity.loc[similarity['OTU_ID'] == corr[1]].iloc[0]['GENOME_ID']
pairs.append((join(params['analysis_folder'], params['model_folder'], '{0}.json'.format(one)),
join(params['analysis_folder'], params['model_folder'], '{0}.json'.format(two))))
# Create two species community models.
try:
cherrypy.log('Widget 4: Started creating community models from {0} pairs of single species models'
.format(len(pairs)))
model_filenames = create_interaction_models(pairs, output_folder=community_folder)
output_filename = join(params['analysis_folder'], params['pair_models_file'])
with open(output_filename, 'w') as handle:
handle.write('\n'.join(model_filenames)+'\n')
cherrypy.log("Widget 4: Finished creating {0} community models".format(len(model_filenames)))
except Exception as e:
cherrypy.log("Widget 4: Error creating community models: {0}".format(e))
return "Sorry something's wrong. Make sure the path to your file is correct and " \
"that the Python package cobrapy is loaded into your system.<br><br>Exception: {0}".format(e)
# Generate the output for the Results tab.
return 'We created {0} two species community models. In the next widget, we will use them ' \
'to predict the growth rate of their species in isolation and when in the community ' \
'using COBRA tools. A list of the community model file names was stored in the ' \
'"{1}" file.'.format(len(model_filenames), output_filename)
def test_bad_extension(self, data_folder, test_folder):
model_files = ['BT.sbml', 'FP.bad']
source_models = mminte.get_all_pairs(
[join(data_folder, x) for x in model_files])
with pytest.raises(IOError):
mminte.create_interaction_models(source_models, test_folder)
def test_not_enough_source(self, data_folder, test_folder):
with pytest.raises(ValueError):
mminte.create_interaction_models([(join(data_folder, 'BT.sbml'))],
test_folder)
def test_bad_extension(self, data_folder):
model_files = ['BT.sbml', 'FP.bad']
source_models = [join(data_folder, x) for x in model_files]
with pytest.raises(IOError):
mminte.create_interaction_models(source_models, output_folder=gettempdir())
