def mapDrugPairs(list, xref):
""" map drug pairs to PDDI dicts. Each key points to a list of PDDIs that share a drug pair """
pddiDictL = []
for p in list:
d1 = p["drug1"]
d2 = p["drug2"]
if xref.has_key(d1) and xref.has_key(d2):
pddi = getPDDIDict()
(pddi["object"], pddi["drug1"], pddi["precipitant"], pddi["drug2"],
pddi["severity"], pddi["uri"], pddi["label"],
pddi["source"]) = (xref[d1]["drugname"],
"http://bio2rdf.org/drugbank:" +
str(xref[d1]["mappedIDs"][0]),
xref[d2]["drugname"],
"http://bio2rdf.org/drugbank:" +
str(xref[d2]["mappedIDs"][0]), p["severity"],
p["uri"], p["label"], "NDF-RT")
print "object: %s: %s-- precipitant:%s : %s" % (
pddi["object"], pddi["drug1"], pddi["precipitant"],
pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def mapDrugPairs(list, xref):
""" map drug pairs to PDDI dicts. Each key points to a list of PDDIs that share a drug pair """
pddiDictL = []
for p in list:
d1 = p["drug1"]
d2 = p["drug2"]
if xref.has_key(d1) and xref.has_key(d2):
pddi = getPDDIDict()
(
pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["severity"],
pddi["uri"],
pddi["label"],
pddi["source"]
)=(
xref[d1]["drugname"],
"http://bio2rdf.org/drugbank:" + str(xref[d1]["mappedIDs"][0]),
xref[d2]["drugname"],
"http://bio2rdf.org/drugbank:" + str(xref[d2]["mappedIDs"][0]),
p["severity"],
p["uri"],
p["label"],
"NDF-RT"
)
print "object: %s: %s-- precipitant:%s : %s" % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def WorldVista_Pickle_Generator(datafile):
# open the World-Vista data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# World-Vista PDDI.
pddiDictL = []
f = open(datafile, 'r')
while 1 :
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug_1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(
pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["severity"],
pddi["homepage"],
pddi["ddiPkMechanism"],
pddi["label"],
pddi["source"]
)=(
elts[0],
"http://bio2rdf.org/drugbank:" + str(elts[1]),
elts[2],
"http://bio2rdf.org/drugbank:" +str(elts[3]),
elts[4],
elts[5],
elts[6],
elts[7],
"World-Vista"
)
#print "ID: %s - object: %s: %s-- precipitant:%s : %s" % (elts[0],pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
print "object: %s: %s-- precipitant:%s : %s" % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def NLMCorpus_Pickle_Generator(datafile):
# open the NLMCorpus_ data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# NLMCorpus_Mapped PDDI.
pddiDictL = []
f = open(datafile, "r")
while 1:
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug_1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(
pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["ddiType"],
pddi["evidenceStatement"],
pddi["source"],
) = (
elts[0],
"http://bio2rdf.org/drugbank:" + str(elts[2]),
elts[3],
"http://bio2rdf.org/drugbank:" + str(elts[5]),
elts[6],
elts[7],
"NLM-Corpus",
)
print "object: %s: %s-- precipitant:%s : %s" % (
pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
)
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def OSCAR_Pickle_Generator():
# open the OSCAR data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# OSCAR PDDI.
pddiDictL = []
f = open(OSCAR_DATA, 'r')
while 1 :
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug_1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(
pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["effectConcept"],
pddi["severity"],
pddi["evidence"],
pddi["evidenceStatement"],
pddi["source"]
)=(
elts[1],
elts[3],
elts[5],
elts[7],
elts[8],
elts[9],
elts[10],
elts[11],
"OSCAR"
)
print "object: %s: %s-- precipitant:%s : %s" % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def PKCorpus_Pickle_Generator(datafile):
# open the PKCorpus_ data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# PKCorpus_Mapped PDDI.
pddiDictL = []
f = open(datafile, 'r')
while 1 :
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug_1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(
pddi["object"],
pddi["drug1"],
pddi["whoAnnotated"],
pddi["precipitant"],
pddi["drug2"],
#pddi["ddiPkEffect"],
pddi["evidenceStatement"],
pddi["source"]
)=(
elts[0],
"http://bio2rdf.org/drugbank:" + str(elts[3]),
elts[6],
elts[4],
"http://bio2rdf.org/drugbank:" +str(elts[7]),
#elts[10],
elts[11],
"PK-Corpus"
)
print "object: %s: %s-- precipitant:%s : %s" % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def hiv_DDI_Pickle_Generator():
# open the HIV-drug-interactions data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# HIV-drug-interactions .
pddiDictL = []
f = open(HIV_DDI_DATA, 'r')
while 1 :
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug 1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("\t")
pddi = getPDDIDict()
(pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["evidenceStatement"],
pddi["label"],
pddi["source"]
) = (
elts[0],
"http://bio2rdf.org/drugbank:" + str(elts[1]),
elts[2],
"http://bio2rdf.org/drugbank:" + str(elts[3]),
elts[4],
elts[5],
"HIV"
)
print "object: %s: %s-- precipitant:%s : %s" % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
#print "effectConcept: %s:-- ddiPkMechanism: %s" % (pddi["effectConcept"],pddi["ddiPkMechanism"] )
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def DDICorpus2011_Pickle_Generator(datafile):
# open the DDI-Corpus2011 data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# DDI-Corpus2011 PDDI.
pddiDictL = []
f = open(datafile, 'r')
while 1 :
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("object") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["evidenceStatement"],
pddi["source"]
) = (
elts[0],
"http://bio2rdf.org/drugbank:" + str(elts[1]),
elts[2],
"http://bio2rdf.org/drugbank:" + str(elts[3]),
elts[4],
"DDI-Corpus-2011"
)
print "object: %s: %s-- precipitant:%s : %s " % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
#print "ID: %s - object: %s: %s-- precipitant:%s : %s -evidence: %s " % (elts[0],pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"], pddi["evidenceStatement"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def Drugbank4_Pickle_Generator():
# open the load-Drugbank4-DDIs data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# a PDDI.
pddiDictL = []
f = open(Drugbank4_DATA, 'r')
while 1 :
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("object_drugbank") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
#elts = l.strip("\n").split(",")
pddi = getPDDIDict()
(pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["source"]
) = (
elts[1],
"http://bio2rdf.org/drugbank:" + str(elts[0]),
elts[3],
"http://bio2rdf.org/drugbank:" + str(elts[2]),
"Drugbank"
)
#print "object: %s -- precipitant:%s " %(pddi["object"],pddi["precipitant"])
print "object: %s: %s-- precipitant:%s : %s" % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def Drugbank_Pickle_Generator():
# open the load-Drugbank4-DDIs data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# a PDDI.
pddiDictL = []
f = open(Drugbank_DATA, 'r')
while 1 :
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("object_drugbank") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["label"],
pddi["source"]
) = (
elts[1],
"http://bio2rdf.org/drugbank:" + str(elts[0]),
elts[3],
"http://bio2rdf.org/drugbank:" + str(elts[2]),
elts[4],
"Drugbank"
)
#print "object: %s: %s-- precipitant:%s : %s" % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def frenchDB_Pickle_Generator():
# open the FRENCHDB data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# FRENCHDB PDDI.
# Dataset contains fields:"mol2"(drug2),"mol1"(drug1),"prota2"(class2),"prota1"(class1),
# "description_interaction","mecanisme","niveau"(severity level),
# "DB1"(drug1 Drugbank ID),"DB2"(drug2 Drugbank ID)
pddiDictL = []
f = open(FRENCHDB_DATA, 'rb')
reader = csv.reader(f, delimiter='\t', quotechar='|')
while 1:
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("mol1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("\t")
pddi = getPDDIDict()
(pddi["object"], pddi["drug1"], pddi["precipitant"], pddi["drug2"],
pddi["label"], pddi["severity"],
pddi["source"]) = (elts[0].strip("\""),
"http://bio2rdf.org/drugbank:" +
str(elts[8]).strip("\""), elts[1].strip("\""),
"http://bio2rdf.org/drugbank:" +
str(elts[7]).strip("\""), elts[4].strip("\""),
elts[6].strip("\""), "FrenchDB")
print "object: %s: %s-- precipitant:%s : %s" % (
pddi["object"], pddi["drug1"], pddi["precipitant"],
pddi["drug2"])
#print "effectConcept: %s:-- ddiPkMechanism: %s" % (pddi["effectConcept"],pddi["ddiPkMechanism"] )
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def ONC_NonInteruptive_Pickle_Generator():
# open the ONCHighPriority data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# ONCHighPriority PDDI.
pddiDictL = []
f = open(ONCNonInteruptive_DATA, 'r')
while 1 :
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug_1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(
pddi["object"],
pddi["drug1"],
pddi["precipitant"],
pddi["drug2"],
pddi["source"]
)=(
elts[0],
"http://bio2rdf.org/drugbank:" + str(elts[1]),
elts[2],
"http://bio2rdf.org/drugbank:" + str(elts[3]),
"ONC-NonInteruptive"
)
print "object: %s: %s-- precipitant:%s : %s" % (pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def DDICorpus2011_Pickle_Generator(datafile):
# open the DDI-Corpus2011 data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# DDI-Corpus2011 PDDI.
pddiDictL = []
f = open(datafile, 'r')
while 1:
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("object") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(pddi["object"], pddi["drug1"], pddi["precipitant"], pddi["drug2"],
pddi["evidenceStatement"],
pddi["source"]) = (elts[0], "http://bio2rdf.org/drugbank:" +
str(elts[1]), elts[2],
"http://bio2rdf.org/drugbank:" + str(elts[3]),
elts[4], "DDI-Corpus-2011")
print "object: %s: %s-- precipitant:%s : %s " % (
pddi["object"], pddi["drug1"], pddi["precipitant"],
pddi["drug2"])
#print "ID: %s - object: %s: %s-- precipitant:%s : %s -evidence: %s " % (elts[0],pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"], pddi["evidenceStatement"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def WorldVista_Pickle_Generator(datafile):
# open the World-Vista data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# World-Vista PDDI.
pddiDictL = []
f = open(datafile, 'r')
while 1:
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug_1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(pddi["object"], pddi["drug1"], pddi["precipitant"], pddi["drug2"],
pddi["severity"], pddi["homepage"], pddi["ddiPkMechanism"],
pddi["label"],
pddi["source"]) = (elts[0], "http://bio2rdf.org/drugbank:" +
str(elts[1]), elts[2],
"http://bio2rdf.org/drugbank:" + str(elts[3]),
elts[4], elts[5], elts[6], elts[7], "World-Vista")
#print "ID: %s - object: %s: %s-- precipitant:%s : %s" % (elts[0],pddi["object"],pddi["drug1"],pddi["precipitant"], pddi["drug2"])
print "object: %s: %s-- precipitant:%s : %s" % (
pddi["object"], pddi["drug1"], pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def hiv_DDI_Pickle_Generator():
# open the HIV-drug-interactions data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# HIV-drug-interactions .
pddiDictL = []
f = open(HIV_DDI_DATA, 'r')
while 1:
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug 1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("\t")
pddi = getPDDIDict()
(pddi["object"], pddi["drug1"], pddi["precipitant"], pddi["drug2"],
pddi["evidenceStatement"], pddi["label"],
pddi["source"]) = (elts[0], "http://bio2rdf.org/drugbank:" +
str(elts[1]), elts[2],
"http://bio2rdf.org/drugbank:" + str(elts[3]),
elts[4], elts[5], "HIV")
print "object: %s: %s-- precipitant:%s : %s" % (
pddi["object"], pddi["drug1"], pddi["precipitant"],
pddi["drug2"])
#print "effectConcept: %s:-- ddiPkMechanism: %s" % (pddi["effectConcept"],pddi["ddiPkMechanism"] )
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
def twosides_f_generator():
# open the twosides data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# twosides PDDI.
f = open(TWOSIDES_DATA, 'r')
while 1:
if not f:
f.close()
raise StopIteration
l = f.readline()
while l.find("CID") == -1: # skips header and stops at EOF
l = f.readline()
if l == "":
f.close()
raise StopIteration
elts = l.strip("\n").split("\t")
pddi = getPDDIDict()
(pddi["drug1"],
pddi["drug2"],
#pddi["label"],
pddi["effectConcept"],
pddi["certainty"],
pddi["source"]
) = (
elts[0],
elts[1],
#elts[4],
elts[5],
elts[7],
"Twosides"
)
# return the dictionary entry
yield pddi
def OSCAR_Pickle_Generator():
# open the OSCAR data file and parse it incrementally
#
# @returns: a PDDI drug model dictionary containing all of the data in a single
# OSCAR PDDI.
pddiDictL = []
f = open(OSCAR_DATA, 'r')
while 1:
if not f:
f.close()
raise StopIteration
l = f.readline()
if l == "":
f.close()
break
else:
if l.find("Drug_1") != -1: # skips header and stops at EOF
l = f.readline()
elts = l.strip("\n").split("$")
pddi = getPDDIDict()
(pddi["object"], pddi["drug1"], pddi["precipitant"], pddi["drug2"],
pddi["effectConcept"], pddi["severity"], pddi["evidence"],
pddi["evidenceStatement"],
pddi["source"]) = (elts[1], elts[3], elts[5], elts[7], elts[8],
elts[9], elts[10], elts[11], "OSCAR")
print "object: %s: %s-- precipitant:%s : %s" % (
pddi["object"], pddi["drug1"], pddi["precipitant"], pddi["drug2"])
# return the dictionary entry
pddiDictL.append(pddi)
return pddiDictL
print "OBSERVED DDIs query_string: %s" % query_string
json_string = query(query_string, sparql_service)
resultset=json.loads(json_string)
#print resultset.values()
print str(len(resultset["results"]["bindings"]))
if len(resultset["results"]["bindings"]) == 0:
print "INFO: No result!"
else:
#print json.dumps(resultset,indent=1)
for i in range(0, len(resultset["results"]["bindings"])):
#for i in range(10, 11):
newPDDI = getPDDIDict()
newPDDI["evidence"] = resultset["results"]["bindings"][i]["evidence"]["value"].encode("utf8")
newPDDI["researchStatement"] = resultset["results"]["bindings"][i]["asrt"]["value"].encode("utf8")
newPDDI["uri"] = resultset["results"]["bindings"][i]["s"]["value"].encode("utf8")
obj = resultset["results"]["bindings"][i]["object"]["value"].encode("utf8")
newPDDI["object"] = obj.replace(u"http://dbmi-icode-01.dbmi.pitt.edu/dikb/resource/Drugs/",u"").upper().encode("utf8")
precip = resultset["results"]["bindings"][i]["precip"]["value"].encode("utf8")
newPDDI["precip"] = precip.replace(u"http://dbmi-icode-01.dbmi.pitt.edu/dikb/resource/Drugs/",u"").upper().encode("utf8")
newPDDI["objectURI"] = resultset["results"]["bindings"][i]["objectURI"]["value"].encode("utf8")
newPDDI["precipURI"] = resultset["results"]["bindings"][i]["precipURI"]["value"].encode("utf8")
newPDDI["label"] = resultset["results"]["bindings"][i]["label"]["value"].encode("utf8")
newPDDI["homepage"] = resultset["results"]["bindings"][i]["homepage"]["value"].encode("utf8")
newPDDI["numericVal"] = resultset["results"]["bindings"][i]["numericVal"]["value"].encode("utf8")
skippedCnt += 1
continue
# test for duplicate
k1 = "%s-%s" % (drug1, drug2)
k2 = "%s-%s" % (drug2, drug1)
if keyDict.get(k1) != None or keyDict.get(k2) != None:
print "INFO: duplicate PDDI (%s) not adding to the list" % k1
dupCnt += 1
continue
keyDict[k1] = None
keyDict[k2] = None
d1 = keggIdToDrugBankD.get(drug1).split("$")
d2 = keggIdToDrugBankD.get(drug2).split("$")
newPDDI = getPDDIDict()
newPDDI["source"] = keggAPIURL
newPDDI["uri"] = keggAPIURL + kd
newPDDI["drug1"] = d1[0]
newPDDI["object"] = d1[1]
newPDDI["drug2"] = d2[0]
newPDDI["precipitant"] = d2[1]
newPDDI["source"] = "Kegg"
if p_or_c == "P":
newPDDI["precaution"] = True
else:
newPDDI["contraindication"] = True
if pkMech == "unclassified":
newPDDI["ddiPkMechanism"] = pkMech