def sbml_to_padmetRef(sbml, padmetRef_file, output=None, db="NA", version="NA", verbose=False):
"""
if padmetRef, not padmetSpec:
if padmetRef exist, instance PadmetRef
else init PadmetRef
update padmetRef
if padmetSpec:
if padmetRef, check if exist else raise Error
if padmetSpec exist, instance PadmetSpec
else init PadmetSpec
update padmetSpec using padmetRef if padmetRef
#TODO
"""
if output is None:
output = padmetRef_file
if os.path.isdir(sbml):
sbml_files = [os.path.join(sbml,_f) for _f in next(os.walk(sbml))[2] if _f.endswith(".sbml") or _f.endswith(".xml")]
else:
sbml_files = sbml.split(";")
if os.path.isfile(padmetRef_file):
padmet_to_update = PadmetRef(padmetRef_file)
else:
padmet_to_update = create_padmet_instance(padmetRef_file, "PadmetRef", db, version)
for sbml_file in sbml_files:
if verbose:
print("Updating padmet from %s" %os.path.basename(sbml_file))
padmet_to_update.updateFromSbml(sbml_file, verbose)
padmet_to_update.generateFile(output)
def biggAPI_to_padmet(output, pwy_file=None, verbose=False):
"""
Extract BIGG database using the api. Create a padmet file.
Escape reactions of biomass.
Require internet access !
Allows to extract the bigg database from the API to create a padmet.
1./ Get all reactions universal id from http://bigg.ucsd.edu/api/v2/universal/reactions, escape reactions of biomass.
2./ Using async_list, extract all the informations for each reactions (compounds, stochio, name ...)
3./ Need to use sleep time to avoid to lose the server access.
4./ Because the direction fo the reaction is not set by default in bigg.
We get all the models where the reaction is and the final direction will the one found
in more than 75%
5./ Also extract xrefs
Parameters
----------
output: str
path to output, the padmet file.
pwy_file: str
path to pathway file, add kegg pathways, line:'pwy_id, pwy_name, x, rxn_id'.
verbose: bool
if True print information
"""
now = datetime.now()
today_date = now.strftime("%Y-%m-%d")
#print(verbose,today_date,version, output, classes_file, compounds_file, proteins_file, reactions_file, enzrxns_file, pathways_file)
policyInArray = [
['compound', 'has_name', 'name'], ['compound', 'has_xref', 'xref'],
['compound', 'has_suppData', 'suppData'], ['gene', 'has_name', 'name'],
['gene', 'has_xref', 'xref'], ['gene', 'has_suppData', 'suppData'],
['gene', 'codes_for', 'protein'], ['pathway', 'has_name', 'name'],
['pathway',
'has_xref', 'xref'], ['pathway', 'is_in_pathway', 'pathway'],
['protein', 'has_name', 'name'], ['protein', 'has_xref', 'xref'],
['protein', 'has_suppData',
'suppData'], ['protein', 'catalyses', 'reaction'],
['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": "BIGG",
"version": "1.5"
}
}
padmetRef = PadmetRef()
if verbose: print("setting policy")
padmetRef.setPolicy(policyInArray)
if verbose: print("setting dbInfo")
padmetRef.setInfo(dbNotes)
list_of_relation = []
if verbose: print("Getting all reactions ids")
url_bigg = 'http://bigg.ucsd.edu/api/v2/'
raw_data = requests.get(url_bigg + "universal/reactions").json()['results']
all_reactions_ids = [
rxn_dict['bigg_id'] for rxn_dict in raw_data
if not rxn_dict['bigg_id'].startswith("BIOMASS")
]
if verbose: print("%s reactions to extract" % (len(all_reactions_ids)))
"""
if verbose: print("Extracting informations... Wait")
step = 100
rxn_lower_index = -(step)
rxn_upper_index = 0
rxn_responses = []
all_range = len(all_reactions_ids)/step
for i in range(all_range):
async_list = []
rxn_lower_index += step
rxn_upper_index += step
for rxn_id in all_reactions_ids[rxn_lower_index:rxn_upper_index]:
action_item = grequests.get(url_bigg + "universal/reactions/" +rxn_id)
async_list.append(action_item)
new_responses = [r.json() for r in grequests.map(async_list)]
rxn_responses += new_responses
print("%s/%s done" %(len(rxn_responses),len(all_reactions_ids)))
if rxn_upper_index != len(all_reactions_ids):
async_list = []
last_index = len(all_reactions_ids) - rxn_upper_index
rxn_lower_index += step
rxn_upper_index += last_index
for rxn_id in all_reactions_ids[rxn_lower_index:rxn_upper_index]:
action_item = grequests.get(url_bigg + "universal/reactions/" +rxn_id)
async_list.append(action_item)
new_responses = [r.json() for r in grequests.map(async_list)]
rxn_responses += new_responses
"""
if verbose: print("updating padmet")
count = 0
all_reactions_ids = [
i for i in all_reactions_ids if 'biomass' not in i.upper()
]
for rxn_id in [
i for i in all_reactions_ids if not i.startswith("BIOMASS")
]:
count += 1
if verbose:
print("reaction: %s, %s/%s" %
(rxn_id, count, len(all_reactions_ids)))
if rxn_id not in list(padmetRef.dicOfNode.keys()):
rxn_response = requests.get(url_bigg + "universal/reactions/" +
rxn_id)
rxn_dict = rxn_response.json()
rxn_metabolites = rxn_dict["metabolites"]
if len(rxn_metabolites) > 1:
rxn_id = rxn_dict['bigg_id']
rxn_name = rxn_dict["name"]
all_models_id = [
i["bigg_id"]
for i in rxn_dict["models_containing_reaction"]
]
async_list = []
for model_id in all_models_id:
action_item = grequests.get(url_bigg + "models/" +
model_id + "/reactions/" +
rxn_id)
async_list.append(action_item)
models_responses = [
r.json() for r in grequests.map(async_list)
]
all_lower_bound = [
i["results"][0]["lower_bound"] for i in models_responses
]
ratio_not_rev = float(all_lower_bound.count(0)) / float(
len(all_lower_bound))
if verbose:
print("Reaction not reversible in %s/%s model(s)" %
(all_lower_bound.count(0), len(all_lower_bound)))
if ratio_not_rev >= 0.75:
rxn_direction = "LEFT-TO-RIGHT"
if verbose: print("Reaction not reversible")
else:
rxn_direction = "REVERSIBLE"
if verbose: print("Reaction reversible")
padmetRef.createNode("reaction", rxn_id, {
"COMMON_NAME": [rxn_name],
"DIRECTION": [rxn_direction]
})
rxn_xrefs = rxn_dict["database_links"]
xref_id = rxn_id + "_xrefs"
xref_node = padmetRef.createNode("xref", xref_id)
has_xref_rlt = Relation(rxn_id, "has_xref", xref_id)
list_of_relation.append(has_xref_rlt)
for db, k in list(rxn_xrefs.items()):
_id = k[0]["id"]
if db in list(xref_node.misc.keys()
) and _id not in xref_node.misc[db]:
xref_node.misc[db].append(_id)
else:
xref_node.misc[db] = [_id]
for metabo_dict in rxn_metabolites:
metabo_id = metabo_dict["bigg_id"]
metabo_name = metabo_dict["name"]
metabo_compart = metabo_dict["compartment_bigg_id"]
metabo_stoich = metabo_dict["stoichiometry"]
try:
padmetRef.dicOfNode[metabo_id]
except KeyError:
padmetRef.createNode("compound", metabo_id,
{"COMMON_NAME": [metabo_name]})
if metabo_stoich < 0:
consumes_rlt = Relation(
rxn_id, "consumes", metabo_id, {
"STOICHIOMETRY": [abs(metabo_stoich)],
"COMPARTMENT": [metabo_compart]
})
list_of_relation.append(consumes_rlt)
else:
produces_rlt = Relation(
rxn_id, "produces", metabo_id, {
"STOICHIOMETRY": [abs(metabo_stoich)],
"COMPARTMENT": [metabo_compart]
})
list_of_relation.append(produces_rlt)
else:
if verbose: print("%s already in padmet" % rxn_id)
continue
if verbose: print("Adding all relations")
count = 0
for rlt in list_of_relation:
count += 1
if verbose: print("relation %s/%s" % (count, len(list_of_relation)))
try:
padmetRef.dicOfRelationIn[rlt.id_in].append(rlt)
except KeyError:
padmetRef.dicOfRelationIn[rlt.id_in] = [rlt]
try:
padmetRef.dicOfRelationOut[rlt.id_out].append(rlt)
except KeyError:
padmetRef.dicOfRelationOut[rlt.id_out] = [rlt]
if pwy_file:
add_kegg_pwy(pwy_file, padmetRef, verbose)
if verbose: print("Generating file: %s" % output)
padmetRef.generateFile(output)
def modelSeed_to_padmet(rxn_file, pwy_file, output, verbose=False):
"""
#TODO
"""
global list_of_relation
now = datetime.now()
today_date = now.strftime("%Y-%m-%d")
#print(verbose,today_date,version, output, classes_file, compounds_file, proteins_file, reactions_file, enzrxns_file, pathways_file)
policyInArray = [
['compound', 'has_name', 'name'], ['compound', 'has_xref', 'xref'],
['compound', 'has_suppData', 'suppData'], ['gene', 'has_name', 'name'],
['gene', 'has_xref', 'xref'], ['gene', 'has_suppData', 'suppData'],
['gene', 'codes_for', 'protein'], ['pathway', 'has_name', 'name'],
['pathway',
'has_xref', 'xref'], ['pathway', 'is_in_pathway', 'pathway'],
['protein', 'has_name', 'name'], ['protein', 'has_xref', 'xref'],
['protein', 'has_suppData',
'suppData'], ['protein', 'catalyses', 'reaction'],
['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": "MODELSEED",
"version": "1.0"
}
}
padmetRef = PadmetRef()
if verbose: print("setting policy")
padmetRef.setPolicy(policyInArray)
if verbose: print("setting dbInfo")
padmetRef.setInfo(dbNotes)
list_of_relation = []
rxn_data = json.load(open(rxn_file))
#remove biomass rxn:
rxn_data.pop("rxn12985")
if verbose: print("updating padmet")
count = 0
for rxn_id, rxn_dict in list(rxn_data.items()):
count += 1
if verbose:
print("reaction: %s, %s/%s" % (rxn_id, count, len(rxn_data)))
try:
if not rxn_dict["compound_ids"]:
raise KeyError
except KeyError:
print(rxn_id)
continue
if rxn_id not in list(padmetRef.dicOfNode.keys()):
if rxn_dict["reversibility"] == ">":
rxn_direction = "LEFT-TO-RIGHT"
else:
rxn_direction = "REVERSIBLE"
rxn_name = rxn_dict["name"]
padmetRef.createNode("reaction", rxn_id, {
"COMMON_NAME": [rxn_name],
"DIRECTION": [rxn_direction]
})
rxn_metabolites = rxn_dict["stoichiometry"].split(";")
for metabo_data in rxn_metabolites:
metabo_data = metabo_data.replace("???", "\"")
try:
metabo_temp, metabo_name = metabo_data.split("\"")[:2]
metabo_stoich, metabo_id, metabo_compart = metabo_temp.split(
":")[:3]
except ValueError:
metabo_stoich, metabo_id, metabo_compart, metabo_name = metabo_data.split(
":")[:4]
metabo_stoich = float(metabo_stoich)
#from modelSeed github
if metabo_compart == "0":
metabo_compart = "c"
elif metabo_compart == "1":
metabo_compart = "e"
elif metabo_compart == "2":
metabo_compart = "p"
try:
padmetRef.dicOfNode[metabo_id]
except KeyError:
padmetRef.createNode("compound", metabo_id,
{"COMMON_NAME": [metabo_name]})
if metabo_stoich < 0:
consumes_rlt = Relation(
rxn_id, "consumes", metabo_id, {
"STOICHIOMETRY": [abs(metabo_stoich)],
"COMPARTMENT": [metabo_compart]
})
list_of_relation.append(consumes_rlt)
else:
produces_rlt = Relation(
rxn_id, "produces", metabo_id, {
"STOICHIOMETRY": [abs(metabo_stoich)],
"COMPARTMENT": [metabo_compart]
})
list_of_relation.append(produces_rlt)
else:
if verbose: print("%s already in padmet" % rxn_id)
continue
with open(pwy_file) as csvfile:
reader = csv.DictReader(csvfile, delimiter='\t')
pwy_raw_data = [row for row in reader]
for pwy_raw in pwy_raw_data:
pwy_id = pwy_raw["Source ID"]
pwy_names = [pwy_raw["Name"], pwy_raw["Aliases"]]
rxn_ids = pwy_raw["Reactions"].split("|")
try:
padmetRef.dicOfNode[pwy_id]
except KeyError:
padmetRef.createNode("pathway", pwy_id, {"COMMON_NAME": pwy_names})
for rxn_id in rxn_ids:
pwy_rlt = Relation(rxn_id, "is_in_pathway", pwy_id)
list_of_relation.append(pwy_rlt)
if verbose: print("Adding all relations")
count = 0
for rlt in list_of_relation:
count += 1
if verbose: print("relation %s/%s" % (count, len(list_of_relation)))
try:
padmetRef.dicOfRelationIn[rlt.id_in].append(rlt)
except KeyError:
padmetRef.dicOfRelationIn[rlt.id_in] = [rlt]
try:
padmetRef.dicOfRelationOut[rlt.id_out].append(rlt)
except KeyError:
padmetRef.dicOfRelationOut[rlt.id_out] = [rlt]
"""
if pwy_file:
add_kegg_pwy(pwy_file, padmetRef, verbose)
"""
if verbose: print("Generating file: %s" % output)
padmetRef.generateFile(output)