def test_compare_padmet_cli():
subprocess.call([
'padmet', 'pgdb_to_padmet', '--pgdb', 'test_data/pgdb', '--output',
'test.padmet', '--extract-gene'
])
fabo_1_padmetSpec = PadmetSpec('test.padmet')
os.remove('test.padmet')
fabo_1_padmetSpec.delNode('ACYLCOASYN-RXN')
fabo_1_padmetSpec.generateFile('fabo_1.padmet')
subprocess.call([
'padmet', 'pgdb_to_padmet', '--pgdb', 'test_data/pgdb', '--output',
'test.padmet', '--extract-gene'
])
fabo_2_padmetSpec = PadmetSpec('test.padmet')
os.remove('test.padmet')
fabo_2_padmetSpec.delNode('ACYLCOADEHYDROG-RXN')
fabo_2_padmetSpec.generateFile('fabo_2.padmet')
subprocess.call([
'padmet', 'compare_padmet', '--padmet', 'fabo_1.padmet,fabo_2.padmet',
'--output', 'output'
])
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')
def from_pgdb_to_padmet(pgdb_folder, db='NA', version='NA', source='GENOME', extract_gene=False, no_orphan=False, enhanced_db=False, padmetRef_file=None, verbose=False):
"""
Parameters
----------
pgdb_folder: str
path to pgdb
db: str
pgdb name, default is 'NA'
version: str
pgdb version, default is 'NA'
source: str
tag reactions for traceability, default is 'GENOME'
extract_gene: bool
if true extract genes information
no_orphan: bool
if true, remove reactions without genes associated
enhanced_db: bool
if true, read metabolix-reactions.xml sbml file and add information in final padmet
padmetRef_file: str
path to padmetRef corresponding to metacyc in padmet format
verbose: bool
if True print information
Returns
-------
padmet.padmetRef:
padmet instance with pgdb within pgdb data
"""
global regex_purge, regex_xref, list_of_relation, def_compart_in, def_compart_out
regex_purge = re.compile("<.*?>|\|")
regex_xref = re.compile('^\((?P<DB>\S*)\s*"(?P<ID>\S*)"')
list_of_relation = []
def_compart_in = "c"
def_compart_out = "e"
#parsing args
source = source.upper()
classes_file, compounds_file, proteins_file, reactions_file, enzrxns_file, pathways_file = \
[os.path.join(pgdb_folder,_file) for _file in ["classes.dat", "compounds.dat", "proteins.dat", "reactions.dat", "enzrxns.dat", "pathways.dat"]]
if enhanced_db:
metabolic_reactions = os.path.join(pgdb_folder,"metabolic-reactions.xml")
else:
metabolic_reactions = None
if extract_gene:
genes_file = os.path.join(pgdb_folder,"genes.dat")
else:
genes_file = None
now = datetime.now()
today_date = now.strftime("%Y-%m-%d")
if padmetRef_file:
padmet = PadmetSpec()
padmetRef = PadmetRef(padmetRef_file)
version = padmetRef.info["DB_info"]["version"]
db = padmetRef.info["DB_info"]["DB"]
dbNotes = {"PADMET":{"creation":today_date,"version":"2.6"},"DB_info":{"DB":db,"version":version}}
padmet.setInfo(dbNotes)
padmet.setPolicy(padmetRef)
with open(reactions_file, 'r') as f:
rxns_id = [line.split(" - ")[1] for line in f.read().splitlines() if line.startswith("UNIQUE-ID")]
count = 0
for rxn_id in rxns_id:
count += 1
if verbose: print("%s/%s Copy %s" %(count, len(rxns_id), rxn_id))
try:
padmet.copyNode(padmetRef, rxn_id)
reconstructionData_id = rxn_id+"_reconstructionData_"+source
if reconstructionData_id in list(padmet.dicOfNode.keys()) and verbose:
print("Warning: The reaction %s seems to be already added from the same source %s" %(rxn_id, source))
reconstructionData = {"SOURCE":[source],"TOOL":["PATHWAYTOOLS"],"CATEGORY":["ANNOTATION"]}
reconstructionData_rlt = Relation(rxn_id,"has_reconstructionData",reconstructionData_id)
padmet.dicOfNode[reconstructionData_id] = Node("reconstructionData", reconstructionData_id, reconstructionData)
padmet._addRelation(reconstructionData_rlt)
except TypeError:
print("%s not in padmetRef" %(rxn_id))
if extract_gene:
if verbose: print("parsing genes")
map_gene_ids = genes_parser(genes_file, padmet, verbose)
if verbose: print("parsing proteins")
dict_protein_gene_id = proteins_parser(proteins_file, padmet, verbose)
mapped_dict_protein_gene_id = map_gene_id(dict_protein_gene_id, map_gene_ids)
if verbose: print("parsing association enzrxns")
enzrxns_parser(enzrxns_file, padmet, mapped_dict_protein_gene_id, source, verbose)
else:
POLICY_IN_ARRAY = [['class','is_a_class','class'], ['class','has_name','name'], ['class','has_xref','xref'], ['class','has_suppData','suppData'],
['compound','is_a_class','class'], ['compound','has_name','name'], ['compound','has_xref','xref'], ['compound','has_suppData','suppData'],
['gene','is_a_class','class'], ['gene','has_name','name'], ['gene','has_xref','xref'], ['gene','has_suppData','suppData'], ['gene','codes_for','protein'],
['pathway','is_a_class','class'], ['pathway','has_name','name'], ['pathway','has_xref','xref'], ['pathway','is_in_pathway','pathway'],
['protein','is_a_class','class'], ['protein','has_name','name'], ['protein','has_xref','xref'], ['protein','has_suppData','suppData'], ['protein','catalyses','reaction'],
['protein','is_in_species','class'],
['reaction','is_a_class','class'], ['reaction','has_name','name'], ['reaction','has_xref','xref'], ['reaction','has_suppData','suppData'], ['reaction','has_reconstructionData','reconstructionData'], ['reaction','is_in_pathway','pathway'],
['reaction','consumes','class','STOICHIOMETRY','X','COMPARTMENT','Y'], ['reaction','produces','class','STOICHIOMETRY','X','COMPARTMENT','Y'],
['reaction','consumes','compound','STOICHIOMETRY','X','COMPARTMENT','Y'], ['reaction','produces','compound','STOICHIOMETRY','X','COMPARTMENT','Y'],
['reaction','consumes','protein','STOICHIOMETRY','X','COMPARTMENT','Y'], ['reaction','produces','protein','STOICHIOMETRY','X','COMPARTMENT','Y'],
['reaction','is_linked_to','gene','SOURCE:ASSIGNMENT','X:Y']]
dbNotes = {"PADMET":{"creation":today_date,"version":"2.6"},"DB_info":{"DB":db,"version":version}}
padmet = PadmetRef()
if verbose: print("setting policy")
padmet.setPolicy(POLICY_IN_ARRAY)
if verbose: print("setting dbInfo")
padmet.setInfo(dbNotes)
if verbose: print("parsing classes")
classes_parser(classes_file, padmet, verbose)
if verbose: print("parsing compounds")
compounds_parser(compounds_file, padmet, verbose)
if verbose: print("parsing reactions")
reactions_parser(reactions_file, padmet, extract_gene, source, verbose)
if verbose: print("parsing pathways")
pathways_parser(pathways_file, padmet, verbose)
if extract_gene:
if verbose: print("parsing genes")
map_gene_ids = genes_parser(genes_file, padmet, verbose)
if verbose: print("parsing proteins")
dict_protein_gene_id = proteins_parser(proteins_file, padmet, verbose)
mapped_dict_protein_gene_id = map_gene_id(dict_protein_gene_id, map_gene_ids)
if verbose: print("parsing association enzrxns")
enzrxns_parser(enzrxns_file, padmet, mapped_dict_protein_gene_id, source, verbose)
if metabolic_reactions is not None:
if verbose: print("enhancing db from metabolic-reactions.xml")
padmet = enhance_db(metabolic_reactions, padmet, extract_gene, verbose)
for rlt in list_of_relation:
try:
padmet.dicOfRelationIn[rlt.id_in].append(rlt)
except KeyError:
padmet.dicOfRelationIn[rlt.id_in] = [rlt]
try:
padmet.dicOfRelationOut[rlt.id_out].append(rlt)
except KeyError:
padmet.dicOfRelationOut[rlt.id_out] = [rlt]
if extract_gene and no_orphan:
all_reactions = [node for node in list(padmet.dicOfNode.values()) if node.type == "reaction"]
rxn_to_del = [r for r in all_reactions if not any([rlt for rlt in padmet.dicOfRelationIn[r.id] if rlt.type == "is_linked_to"])]
for rxn in rxn_to_del:
padmet.delNode(rxn.id)
if verbose:
print("%s/%s orphan reactions (without gene association) deleted" %(len(rxn_to_del), len(all_reactions)))
all_genes_linked = set([rlt.id_out for rlt in padmet.getAllRelation() if rlt.type == "is_linked_to"])
all_genes = set([node.id for node in list(padmet.dicOfNode.values()) if node.type == "gene"])
count = 0
for gene_id in [g for g in all_genes if g not in all_genes_linked]:
count += 1
#if verbose: print("Removing gene without gene assoc %s" %gene_id)
padmet.dicOfNode.pop(gene_id)
if verbose:
print("%s/%s orphan genes (not linked to any reactions) deleted" %(count, len(all_genes)))
rxns = [node.id for node in list(padmet.dicOfNode.values()) if node.type == "reaction"]
for rxn_id in rxns:
cp_rlts = set([rlt.type for rlt in padmet.dicOfRelationIn[rxn_id] if rlt.type in ["consumes","produces"]])
if len(cp_rlts) == 1:
print("rxn only consume or produce, transport ???: %s" %rxn_id)
padmet.delNode(rxn_id)
return padmet