def main():
args = docopt.docopt(__doc__)
output = args["--output"]
verbose = args["-v"]
if args["--padmetRef"]:
padmetRef = PadmetRef(args["--padmetRef"])
else:
padmetRef = None
padmet_path = args["--padmet"]
compare_padmet.compare_padmet(padmet_path, output, padmetRef, verbose)
def analysis_on_group(group_name, groups, config_data, verbose):
"""Create reaction dendrogram and extract specific reactions using metabolic networks.
Args:
group_name (str): Name of the group from group_template.tsv.
groups (list): All the species inside the group.
config_data (dict): Dictionary with all configuration paths.
verbose (bool): Verbose.
"""
database_path = config_data['database_path']
padmet_from_networks_path = config_data['padmet_from_networks_path']
analysis_path = config_data['analysis_path']
all_padmet_path = [
os.path.join(padmet_from_networks_path, name + ".padmet")
for name in groups
]
group_analysis_path = analysis_path + '/' + group_name
if not os.path.isdir(group_analysis_path):
if len(groups) == 1:
sys.exit('A group must contain more than one member.')
for padmet_path in all_padmet_path:
if not os.path.exists(padmet_path):
org_name = os.path.splitext(os.path.basename(padmet_path))[0]
sys.exit(
"Padmet file of organism %s from group %s not found in %s"
% (org_name, group_name, padmet_from_networks_path))
# Compare the padmet to create the reactions.csv file needed to create the reaction dendrogram.
compare_padmet.compare_padmet(padmet_path=",".join(all_padmet_path),
output=group_analysis_path,
padmetRef=database_path,
verbose=verbose)
dendrogram_reactions_distance.reaction_figure_creation(
reaction_file=group_analysis_path + '/reactions.csv',
output=group_analysis_path + '/dendrogram_output',
padmetRef=database_path,
verbose=verbose)
else:
print(
group_analysis_path +
' already exists. Delete it if you want to relaunch the analysis.')
def run_compare(run_id, nb_cpu_to_use, verbose):
"""Compare the gorup specified by the user.
Args:
run_id (str): ID of the run
nb_cpu_to_use (int): number of CPU for multiprocessing
verbose (boolean): verbose
"""
if verbose:
print('--- Running compare step ---')
compare_start_time = time.time()
config_data = parse_config_file(run_id)
analysis_path = config_data['analysis_path']
analysis_group_file_path = config_data['analysis_group_file_path']
compare_output_path = analysis_path + '/compare_group'
database_path = config_data['database_path']
padmet_from_networks_path = config_data['padmet_from_networks_path']
# Create a dictionary containing the group name and the species inside the group.
group_data = {}
padmets = []
with open(analysis_group_file_path, 'r') as group_file:
group_reader = csv.reader(group_file, delimiter='\t')
cluster_reactions = {}
for row in group_reader:
group_name = row[0]
groups = [species for species in row[1:] if species != '']
group_data[group_name] = groups
if group_name != 'all':
padmets.extend([
padmet_from_networks_path + '/' + species + '.padmet'
for species in groups
])
padmets = list(set(padmets))
if not os.path.isdir(compare_output_path):
os.mkdir(compare_output_path)
padmetref = PadmetRef(database_path)
# Create the reactions.tsv file needed to create dendrogram.
padmet_path = ','.join(padmets)
compare_padmet.compare_padmet(padmet_path=padmet_path,
output=compare_output_path,
padmetRef=padmetref,
verbose=verbose)
# Read the reactions.tsv file and remove the column unused.
reactions_file = compare_output_path + '/' + 'reactions.tsv'
reactions_dataframe = pa.read_csv(reactions_file, sep='\t')
columns = [
column for column in reactions_dataframe.columns
if '(sep=;)' not in column and '_formula' not in column
]
reactions_dataframe = reactions_dataframe[columns].copy()
reactions_dataframe.set_index('reaction', inplace=True)
# For each group, extract the reactions present in its species to create supervenn sets.
supervenn_sets = []
supervenn_labels = []
for group_name in group_data:
if group_name != 'all':
groups = group_data[group_name]
reactions_temp = []
for species in groups:
species_reactions_dataframe = reactions_dataframe[
reactions_dataframe[species] == 1]
reactions_temp.extend(
species_reactions_dataframe.index.tolist())
supervenn_sets.append(set(reactions_temp))
supervenn_labels.append(group_name)
cluster_reactions[group_name] = set(reactions_temp)
supervenn(supervenn_sets,
supervenn_labels,
chunks_ordering='occurence',
sets_ordering='minimize gaps')
plt.savefig(compare_output_path + '/compare_group.png',
bbox_inches='tight')
plt.clf()
dendrogram_reactions_distance.reaction_figure_creation(
reactions_file,
os.path.join(compare_output_path, "dendrogram_output"),
padmetRef_file=database_path,
verbose=verbose)
compare_end_time = (time.time() - compare_start_time)
integer_part, decimal_part = str(compare_end_time).split('.')
compare_time = ".".join([integer_part, decimal_part[:3]])
if verbose:
print("--- compare step done in: %ss ---" % compare_time)
def test_compare_padmet():
fabo_1_padmetSpec = from_pgdb_to_padmet('test_data/pgdb',
extract_gene=True)
fabo_1_padmetSpec.delNode('ACYLCOASYN-RXN')
fabo_1_padmetSpec.generateFile('fabo_1.padmet')
fabo_2_padmetSpec = from_pgdb_to_padmet('test_data/pgdb',
extract_gene=True)
fabo_2_padmetSpec.delNode('ACYLCOADEHYDROG-RXN')
fabo_2_padmetSpec.generateFile('fabo_2.padmet')
compare_padmet('fabo_1.padmet,fabo_2.padmet',
'output',
padmetRef=None,
verbose=False)
genes_fabo_1 = []
genes_fabo_2 = []
with open('output/genes.tsv', 'r') as genes_file:
csvreader = csv.reader(genes_file, delimiter='\t')
for row in csvreader:
if row[1] == '1':
genes_fabo_1.append(row[0])
if row[2] == '1':
genes_fabo_2.append(row[0])
assert set(FABO_GENES).issubset(set(genes_fabo_1))
assert set(FABO_GENES).issubset(set(genes_fabo_2))
reactions_fabo_1 = []
reactions_fabo_2 = []
with open('output/reactions.tsv', 'r') as reactions_file:
csvreader = csv.reader(reactions_file, delimiter='\t')
for row in csvreader:
if row[1] == '1':
reactions_fabo_1.append(row[0])
if row[2] == '1':
reactions_fabo_2.append(row[0])
expected_fabo_1_rxns = [
rxn for rxn in FABO_RXNS if rxn != 'ACYLCOASYN-RXN'
]
expected_fabo_2_rxns = [
rxn for rxn in FABO_RXNS if rxn != 'ACYLCOADEHYDROG-RXN'
]
assert set(expected_fabo_1_rxns).issubset(set(reactions_fabo_1))
assert set(expected_fabo_2_rxns).issubset(set(reactions_fabo_2))
pathway_fabo_1 = []
pathway_fabo_2 = []
with open('output/pathways.tsv', 'r') as pathways_file:
csvreader = csv.reader(pathways_file, delimiter='\t')
for row in csvreader:
if row[0] != 'pathway':
pathway_fabo_1.append(row[0])
pathway_fabo_2.append(row[0])
if row[3] != 'fabo_1_rxn_assoc (sep=;)':
pwy_reactions_fabo_1 = row[3].split(';')
if row[4] != 'fabo_2_rxn_assoc (sep=;)':
pwy_reactions_fabo_2 = row[4].split(';')
assert pathway_fabo_1 == ['FAO-PWY']
assert pathway_fabo_2 == ['FAO-PWY']
assert set(expected_fabo_1_rxns).issubset(set(pwy_reactions_fabo_1))
assert set(expected_fabo_2_rxns).issubset(set(pwy_reactions_fabo_2))
metabolites_fabo_1 = []
metabolites_fabo_2 = []
with open('output/metabolites.tsv', 'r') as metabolites_file:
csvreader = csv.reader(metabolites_file, delimiter='\t')
for row in csvreader:
if row[1] != 'fabo_1_rxn_consume' or row[1] != '':
if row[0] != 'metabolite':
metabolites_fabo_1.append(row[0])
if row[3] != 'fabo_1_rxn_produce' or row[3] != '':
if row[0] != 'metabolite':
metabolites_fabo_1.append(row[0])
if row[2] != 'fabo_2_rxn_consume' or row[2] != '':
if row[0] != 'metabolite':
metabolites_fabo_2.append(row[0])
if row[2] != 'fabo_2_rxn_produce' or row[2] != '':
if row[0] != 'metabolite':
metabolites_fabo_2.append(row[0])
metabolites_fabo_1 = list(set(metabolites_fabo_1))
metabolites_fabo_2 = list(set(metabolites_fabo_2))
assert set(FABO_CPDS).issubset(set(metabolites_fabo_1))
assert set(FABO_CPDS).issubset(set(metabolites_fabo_2))
os.remove('fabo_1.padmet')
os.remove('fabo_2.padmet')
shutil.rmtree('output')