def __init__(self, drug_decode):
print("ChemSpiderSearch is still in progress, please do not use")
self.dd = DrugDecode(drug_decode)
self.dd_filled = DrugDecode(drug_decode)
from bioservices.chemspider import ChemSpider
from bioservices import ChEMBL
from bioservices import UniChem
try:
print('Loading PubChem')
from bioservices.pubchem import PubChem
self.puchem = PubChem()
except:
# Pubchem was introduced only in dec 2015
pass
print('Loading ChEMBL service')
self.chembl = ChEMBL(cache=True)
print('Loading ChemSpider service')
self.chemspider = ChemSpider(cache=True)
print('Loading UniChem service')
# in unichem db number is 22 and chembl is 1
self.unichem = UniChem()
print('Settings some data aliases')
self._cs_find = self.chemspider.find
self._cs_get = self.chemspider.GetExtendedCompoundInfo
self.drug_ids = sorted(list(self.dd.df.index.values))
self.drug_names = sorted(list(self.dd.df.DRUG_NAME.values))
def get_compound_id(smiles):
""" returns kegg id for compund with given smiles """
indigo = Indigo()
# convert smiles to standard format
mol = indigo.loadMolecule(smiles)
mol.aromatize()
moi_smiles = mol.canonicalSmiles()
# Get list of possible kegg IDs
url = "http://rest.genome.jp/subcomp/?smiles=%s&cutoff=1.0" % smiles
http_client = HTTPClient()
try:
response = http_client.fetch(url).body
except HTTPError as e:
raise RuntimeError("Error:", str(e))
http_client.close()
subcomp_results = response.split("\n")
subcomp_results.pop()
subcomp_results = ([i.split('\t')[0] for i in subcomp_results])
# get smiles for all compound IDs found
all_smiles = []
uni = UniChem()
mapping = uni.get_mapping("kegg_ligand", "chebi")
ch = ChEBI()
all_smiles = [ch.getCompleteEntity(mapping[x]).smiles
for x in subcomp_results]
# convert smiles to a standard format
for pos, mol in enumerate(all_smiles):
m = indigo.loadMolecule(mol)
m.aromatize()
all_smiles[pos] = m.canonicalSmiles()
# check if smiles matches given and, if so, use that compound ID
# if not, errors out
try:
index = all_smiles.index(moi_smiles)
except:
raise RuntimeError("SMILES unmatchable to: %s" % str(all_smiles))
return subcomp_results[index]
def test_get_source_id():
uni = UniChem()
assert uni.get_source_id("chembl") == 1
assert uni.get_source_id("1") == 1
assert uni.get_source_id(1) == 1
try:
uni.get_source_id("wrong")
assert False
except:
assert True
try:
uni.get_source_id("20000")
assert False
except:
assert True
def __init__(self, verbosity="INFO"):
super(Mapper, self).__init__(level=verbosity)
self.logging.info("Initialising the services")
self.logging.info("... uniprots")
self._uniprot_service = UniProt()
self.logging.info("... KEGG")
self._kegg_service = KeggParser(verbose=False)
self.logging.info("... HGNC")
self._hgnc_service = HGNC()
self.logging.info("... UniChem")
self._unichem_service = UniChem()
self.logging.info("...BioDBNet")
self._biodbnet = BioDBNet()
def convert_kegg_nodes(self, network):
"""
Maps network from kegg to gene names
Parameters
----------
network : nx.DiGraph
Returns
-------
dict
"""
chem = UniChem()
still_unknown = []
hits = [i for i in set(network.nodes) if i.startswith('cpd:')]
net_kegg_names = dict()
net_chem_names = dict()
net_cpd_to_hmdb = dict()
for i in hits:
name_stripped = i.lstrip('cpd:')
net_kegg_names[i] = name_stripped
if name_stripped in self.kegg_to_hmdb:
mapping = self.kegg_to_hmdb[name_stripped]
if isinstance(mapping, (list, set, SortedSet)):
names = '|'.join(set(mapping))
chem_names = set()
for name in mapping:
try:
chem_names.update(self.hmdb_to_chem_name[name])
except:
continue
net_cpd_to_hmdb[i] = names
net_chem_names[i] = order_merge(chem_names)
elif isinstance(mapping, basestring):
chem_n = self.hmdb_to_chem_name[mapping]
net_cpd_to_hmdb[i] = mapping
net_chem_names[i] = '|'.join(chem_n.encode('ascii',
'ignore'))
else:
print('Returned something else...', mapping)
elif i in compound_manual:
loc = compound_manual[i]
net_cpd_to_hmdb[i] = loc
if loc in self.hmdb_to_chem_name:
net_chem_names[i] = order_merge(
self.hmdb_to_chem_name[loc])
else:
still_unknown.append(i)
if len(still_unknown):
kegg_hmdb = chem.get_mapping("kegg_ligand", "hmdb")
for i in still_unknown:
name_stripped = i.lstrip('cpd:')
if name_stripped in kegg_hmdb:
net_cpd_to_hmdb[i] = kegg_hmdb[name_stripped]
# else:
# print("Cannot find a HMDB mapping for %s " % i)
return net_cpd_to_hmdb, net_kegg_names, net_chem_names
"""
from argparse import ArgumentParser
import mysql.connector as mysqlc
from bioservices import UniChem
#Define command line options and defaults
parser = ArgumentParser()
parser.add_argument("-c",
"--mysqlconf",
dest="mysql_conf",
default="/home/kkmattil/Documents/DDCB/mysql_write.conf",
help="MySQL_conf",
metavar="MYSQL_CONF")
args = parser.parse_args()
uniC = UniChem()
#Open mysql connection
cnx = mysqlc.connect(option_files=args.mysql_conf)
cursor = cnx.cursor()
SQL_drop_table = ('DROP TABLE unichem_links;')
cursor.execute(SQL_drop_table)
cnx.commit()
cursor = cnx.cursor()
SQL_create_table_if_needed = (
'CREATE TABLE IF NOT EXISTS unichem_links(comp_num INT, source_id INT, source VARCHAR(200), id_in_db VARCHAR(200)) ENGINE=INNODB;'
)
cursor.execute(SQL_create_table_if_needed)
def test_get_auxiliary_mapping():
# this does nothing (behaviour of the function)
uni = UniChem()
res = uni.get_auxiliary_mappings(1)
def test_get_src_compoundid_url():
uni = UniChem()
uni.get_src_compound_id_url("CHEMBL12", "chembl", "drugbank")
def test_unichem_src_compound_from_inchikey():
uni = UniChem()
uni.get_src_compound_ids_from_inchikey("AAOVKJBEBIDNHE-UHFFFAOYSA-N")
uni.get_src_compound_ids_all_from_inchikey("AAOVKJBEBIDNHE-UHFFFAOYSA-N")
def test_get_src_compound_ids_all_from_obsolete():
uni = UniChem()
res = uni.get_src_compound_ids_all_from_obsolete("DB07699", 2)
res = uni.get_src_compound_ids_all_from_obsolete("DB07699", 2, "chembl")
def test_unichem_src_compound():
uni = UniChem()
uni.get_src_compound_ids_from_src_compound_id("CHEMBL12", "chembl", "chebi")
uni.get_src_compound_ids_all_from_src_compound_id("CHEMBL12", "chembl","drugbank")
def test_structure():
uni = UniChem()
uni.get_structure("CHEMBL12", "chembl")
uni.get_structure_all("CHEMBL12", "chembl")
def test_src_ids():
uni = UniChem()
uni.get_all_src_ids()
uni.get_source_information("chembl")
def test_mapping():
uni = UniChem()
res1 = uni.get_mapping("kegg_ligand", "chembl")
assert len(res1)>0
def unichem():
u = UniChem(verbose=False)
return u
def test_get_verbose_src_compound_ids_fron_inchikey():
uni = UniChem()
uni.get_verbose_src_compound_ids_from_inchikey("GZUITABIAKMVPG-UHFFFAOYSA-N")
def convert_kegg_nodes(self, network):
"""
Maps network from kegg to gene names
Parameters
----------
network : nx.DiGraph
Returns
-------
dict
"""
chem = UniChem()
still_unknown = []
hits = [i for i in set(network.nodes) if i.startswith('cpd:')]
net_kegg_names = dict()
net_chem_names = dict()
net_cpd_to_hmdb = dict()
for i in hits:
name_stripped = i.lstrip('cpd:')
net_kegg_names[i] = name_stripped
if name_stripped in self.kegg_to_hmdb:
mapping = self.kegg_to_hmdb[name_stripped]
if isinstance(mapping, (list, set, SortedSet)):
names = '|'.join(set(mapping))
chem_names = set()
for name in mapping:
try:
chem_names.update(self.hmdb_to_chem_name[name])
except:
continue
net_cpd_to_hmdb[i] = names
net_chem_names[i] = order_merge(chem_names)
elif isinstance(mapping, basestring):
chem_n = self.hmdb_to_chem_name[mapping]
net_cpd_to_hmdb[i] = mapping
net_chem_names[i] = '|'.join(
chem_n.encode('ascii', 'ignore'))
else:
print('Returned something else...', mapping)
elif i in compound_manual:
loc = compound_manual[i]
net_cpd_to_hmdb[i] = loc
if loc in self.hmdb_to_chem_name:
net_chem_names[i] = order_merge(
self.hmdb_to_chem_name[loc])
else:
still_unknown.append(i)
if len(still_unknown):
kegg_hmdb = chem.get_mapping("kegg_ligand", "hmdb")
for i in still_unknown:
name_stripped = i.lstrip('cpd:')
if name_stripped in kegg_hmdb:
net_cpd_to_hmdb[i] = kegg_hmdb[name_stripped]
# else:
# print("Cannot find a HMDB mapping for %s " % i)
return net_cpd_to_hmdb, net_kegg_names, net_chem_names