def GetCompoundInfo(self, mm, csid):
"""
take an csid, get a mask
parameters:
-`mm`: a metmask database
-`csid`: a chemspider identifier
"""
tmpmask = mask({}, mm.idpatterns)
csid = self.parent.urlSafe(csid)
url = self.GetCompoundInfoURL + "CSID=" + str(csid) + "&token=" + self.token
qRes = self.parent.getUrl(url)
if not qRes:
return (tmpmask)
tmpmask.append('chemspider', csid, self.parent.confid, self.parent.sourceid)
searchResults = xml.dom.minidom.parse(qRes)
ids = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('CompoundInfo')))
inchi = list(mmquery.nodecontents(searchResults.getElementsByTagName('InChI')))
if inchi:
for ide in inchi:
tmpmask.append('inchi', ide, \
self.parent.confid, self.parent.sourceid)
smiles = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('SMILES')))
if smiles:
for ide in smiles:
tmpmask.append('smiles', ide, self.parent.confid, self.parent.sourceid)
inchikey = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('InChIKey')))
if inchikey:
for ide in inchikey:
tmpmask.append('inchikey', ide, self.parent.confid, self.parent.sourceid)
return (tmpmask)
def getChebiChildren(self, chebiId):
""" get the is_enantiomer children
"""
url = self.getOntologyChildren + "chebiId=" + chebiId
qRes = self.parent.getUrl(url)
if not qRes:
return ([])
searchResults = xml.dom.minidom.parse(qRes)
res = []
for it in searchResults.getElementsByTagName('ns1:ListElement'):
type = list(query.nodecontents(it.getElementsByTagName("ns1:type")))
if type[0] == 'is enantiomer of':
child = list(query.nodecontents(it.getElementsByTagName("ns1:chebiId")))
res.append(child[0].replace("CHEBI:", ""))
return (res)
def InChItool(self, start, goal, query):
urls = {
'inchikey': {
'chemspider': '/InChIKeyToCSID?inchi_key=',
'inchi': '/InChIKeyToInChI?inchi_key='
},
'inchi': {
'chemspider': '/InChIToCSID?inchi=',
'inchikey': '/InChIToInChIKey?inchi=',
'smiles': '/InChIToSMILES?inchi='
},
'smiles': {
'inchi': '/SMILESToInChI?smiles='
}
}
if start not in urls:
raise Exception("unsupported query")
if goal not in urls[start]:
raise Exception("unsupported query")
query = self.parent.urlSafe(query)
url = self.InChI + urls[start][goal] + query
qRes = self.parent.getUrl(url)
if not qRes:
return ([])
searchResults = xml.dom.minidom.parse(qRes)
ids = list(
mmquery.nodecontents(searchResults.getElementsByTagName('string')))
return (ids)
def getChebiChildren(self, chebiId):
""" get the is_enantiomer children
"""
url = self.getOntologyChildren + "chebiId=" + chebiId
qRes = self.parent.getUrl(url)
if not qRes:
return ([])
searchResults = xml.dom.minidom.parse(qRes)
res = []
for it in searchResults.getElementsByTagName('ns1:ListElement'):
type = list(query.nodecontents(
it.getElementsByTagName("ns1:type")))
if type[0] == 'is enantiomer of':
child = list(
query.nodecontents(it.getElementsByTagName("ns1:chebiId")))
res.append(child[0].replace("CHEBI:", ""))
return (res)
def ExtRefs(self, csid, ds):
"""
take a csid
"""
url = self.ExtRefsURL + "CSID=" + str(csid) + "&datasources=" + ds + "&token=" + self.token
qRes = self.parent.getUrl(url)
if not qRes:
return ([])
searchResults = xml.dom.minidom.parse(qRes)
ids = list(mmquery.nodecontents(searchResults.getElementsByTagName('ext_id')))
return (ids)
def url2ids(self, url):
"""
take an url, get the chebi ids
"""
qRes = self.parent.getUrl(url)
if qRes:
searchResults = xml.dom.minidom.parse(qRes)
ids = list(query.nodecontents(searchResults.getElementsByTagName('ns1:chebiId')))
return [x.replace('CHEBI:', '') for x in ids]
else:
return []
def SimpleSearch(self, query):
"""
take a query string, get the csids
"""
query = self.parent.urlSafe(query)
url = self.SimpleSearchURL + "query=" + str(query) + "&token=" + self.token
qRes = self.parent.getUrl(url)
if not qRes:
return ([])
searchResults = xml.dom.minidom.parse(qRes)
ids = list(mmquery.nodecontents(searchResults.getElementsByTagName('int')))
return (ids)
def ExtRefs(self, csid, ds):
"""
take a csid
"""
url = self.ExtRefsURL + "CSID=" + str(
csid) + "&datasources=" + ds + "&token=" + self.token
qRes = self.parent.getUrl(url)
if not qRes:
return ([])
searchResults = xml.dom.minidom.parse(qRes)
ids = list(
mmquery.nodecontents(searchResults.getElementsByTagName('ext_id')))
return (ids)
def url2ids(self, url):
"""
take an url, get the chebi ids
"""
qRes = self.parent.getUrl(url)
if qRes:
searchResults = xml.dom.minidom.parse(qRes)
ids = list(
query.nodecontents(
searchResults.getElementsByTagName('ns1:chebiId')))
return [x.replace('CHEBI:', '') for x in ids]
else:
return []
def GetCompoundInfo(self, mm, csid):
"""
take an csid, get a mask
parameters:
-`mm`: a metmask database
-`csid`: a chemspider identifier
"""
tmpmask = mask({}, mm.idpatterns)
csid = self.parent.urlSafe(csid)
url = self.GetCompoundInfoURL + "CSID=" + str(
csid) + "&token=" + self.token
qRes = self.parent.getUrl(url)
if not qRes:
return (tmpmask)
tmpmask.append('chemspider', csid, self.parent.confid,
self.parent.sourceid)
searchResults = xml.dom.minidom.parse(qRes)
ids = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('CompoundInfo')))
inchi = list(
mmquery.nodecontents(searchResults.getElementsByTagName('InChI')))
if inchi:
for ide in inchi:
tmpmask.append('inchi', ide, \
self.parent.confid, self.parent.sourceid)
smiles = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('SMILES')))
if smiles:
for ide in smiles:
tmpmask.append('smiles', ide, self.parent.confid,
self.parent.sourceid)
inchikey = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('InChIKey')))
if inchikey:
for ide in inchikey:
tmpmask.append('inchikey', ide, self.parent.confid,
self.parent.sourceid)
return (tmpmask)
def SimpleSearch(self, query):
"""
take a query string, get the csids
"""
query = self.parent.urlSafe(query)
url = self.SimpleSearchURL + "query=" + str(
query) + "&token=" + self.token
qRes = self.parent.getUrl(url)
if not qRes:
return ([])
searchResults = xml.dom.minidom.parse(qRes)
ids = list(
mmquery.nodecontents(searchResults.getElementsByTagName('int')))
return (ids)
def InChItool(self, start, goal, query):
urls = {'inchikey': {'chemspider': '/InChIKeyToCSID?inchi_key=',
'inchi': '/InChIKeyToInChI?inchi_key='},
'inchi': {'chemspider': '/InChIToCSID?inchi=',
'inchikey': '/InChIToInChIKey?inchi=',
'smiles': '/InChIToSMILES?inchi='},
'smiles': {'inchi': '/SMILESToInChI?smiles='}}
if start not in urls:
raise Exception("unsupported query")
if goal not in urls[start]:
raise Exception("unsupported query")
query = self.parent.urlSafe(query)
url = self.InChI + urls[start][goal] + query
qRes = self.parent.getUrl(url)
if not qRes:
return ([])
searchResults = xml.dom.minidom.parse(qRes)
ids = list(mmquery.nodecontents(searchResults.getElementsByTagName('string')))
return (ids)
def pubchem2mask(self, docSum):
""" turn a docsum node in to a mask
"""
un = mask({}, self.parent.mm.idpatterns)
cid = next(mmquery.nodecontents(docSum.getElementsByTagName("Id")))
un.append('cid', cid, self.parent.confid, self.parent.sourceid)
p = re.compile('(<a href[^>]*>)|(</a>)|(ligand)|(,)')
cnf = self.parent.confid
src = self.parent.sourceid
weak = self.parent.mm.confidence['weak']
for item in docSum.getElementsByTagName('Item'):
if len(item.childNodes) == 0:
continue
nVal = item.childNodes[0].nodeValue
if item.getAttribute('Name') == 'SynonymList':
synonyms = list(mmquery.nodecontents(item.childNodes))
for s in synonyms:
s = p.sub('', s.strip())
if 'kegg' in guessTable(s, self.parent.mm.idpatterns):
un.append('kegg', s, cnf, src)
continue
if 'cas' in guessTable(s, self.parent.mm.idpatterns):
un.append('cas', s, cnf, src)
continue
if 'chebi' in guessTable(s, self.parent.mm.idpatterns):
un.append('chebi', s, cnf, src)
continue
if 'inchi' in guessTable(s, self.parent.mm.idpatterns):
un.append('inchi', s, cnf, src)
continue
else:
un.append('synonym', s, weak, src)
elif item.getAttribute('Name') == 'IUPACName':
un.append('iupac', nVal, cnf, src)
elif item.getAttribute('Name') == 'CanonicalSmile':
un.append('smiles', nVal, cnf, src)
elif item.getAttribute('Name') == 'CanonicalSmile':
un.append('smiles', nVal, cnf, src)
elif item.getAttribute('Name') == 'InChIKey':
un.append('inchikey', nVal, cnf, src)
# annotation section
elif item.getAttribute('Name') == 'MolecularFormula':
un.append('formula', nVal, weak, src)
elif item.getAttribute('Name') == 'MolecularWeight':
un.append('weight', nVal, weak, src)
elif item.getAttribute('Name') == 'TotalFormalCharge':
un.append('totalcharge', nVal, weak, src)
elif item.getAttribute('Name') == 'XLogP':
un.append('xlogp', nVal, weak, src)
elif item.getAttribute('Name') == 'XLogP':
un.append('xlogp', nVal, weak, src)
elif item.getAttribute('Name') == 'HydrogenBondDonorCount':
un.append('hbonddonor', nVal, weak, src)
elif item.getAttribute('Name') == 'HydrogenBondAcceptorCount':
un.append('hbondacceptor', nVal, weak, src)
elif item.getAttribute('Name') == 'HeavyAtomCount':
un.append('heavyatom', nVal, weak, src)
elif item.getAttribute('Name') == 'TPSA':
un.append('tpsa', nVal, weak, src)
return (un)
def chebi2mask(self, mm, chebiId):
"""
get a mask containing the info associated with a chebi id
2. use getComplete to fetch the contents of the relevant entries
"""
ba = chebiId
un = mask({}, mm.idpatterns)
qUrl = self.getComplete + "chebiId=" + str(chebiId)
qRes = self.parent.getUrl(qUrl)
if not qRes:
return (mask({}))
searchResults = xml.dom.minidom.parse(qRes)
if not searchResults:
return (mask({}))
if searchResults.getElementsByTagName('ns1:return'):
retList = searchResults.getElementsByTagName('ns1:return')[0]
# found non-existent chebiId
else:
# this would have been reasonable but chebi has problems so that some
# entries exist, but cant be fetched. valid chebi but we cant query
# for it,
# hence, skip completely
# delmask = mask({})
# delmask.append('chebi', chebiId)
# mm.brandish(delmask)
return (mask({}))
# chebiid
if retList.getElementsByTagName("ns1:chebiId"):
chebiId = list(
query.nodecontents(
retList.getElementsByTagName("ns1:chebiId")))[0]
un.append('chebi', chebiId.replace("CHEBI:", ""),
self.parent.confid, self.parent.sourceid)
# smiles
if retList.getElementsByTagName("ns1:smiles"):
un.append(
'smiles',
list(
query.nodecontents(
retList.getElementsByTagName("ns1:smiles")))[0],
self.parent.confid, self.parent.sourceid)
# synonym
if searchResults.getElementsByTagName('ns1:Synonyms'):
syns = searchResults.getElementsByTagName('ns1:Synonyms')[0]
for sy in list(
query.nodecontents(syns.getElementsByTagName("ns1:data"))):
un.append('synonym', sy, self.parent.mm.confidence['weak'], \
self.parent.sourceid)
# inchi
if retList.getElementsByTagName("ns1:inchi"):
un.append(
'inchi',
list(
query.nodecontents(
retList.getElementsByTagName("ns1:inchi")))[0],
self.parent.confid, self.parent.sourceid)
# iupac
if searchResults.getElementsByTagName('ns1:IupacNames'):
syns = searchResults.getElementsByTagName('ns1:IupacNames')[0]
for sy in list(
query.nodecontents(syns.getElementsByTagName("ns1:data"))):
un.append('iupac', sy, self.parent.confid, \
self.parent.sourceid)
# kegg
for ll in searchResults.getElementsByTagName('ns1:DatabaseLinks'):
if list(query.nodecontents(ll.getElementsByTagName(
"ns1:type")))[0] == 'KEGG COMPOUND accession':
for sy in list(
query.nodecontents(
ll.getElementsByTagName("ns1:data"))):
un.append('kegg', sy, self.parent.confid,
self.parent.sourceid)
# cas
for ll in searchResults.getElementsByTagName('ns1:RegistryNumbers'):
if list(query.nodecontents(ll.getElementsByTagName(
"ns1:type")))[0] == 'CAS Registry Number':
for sy in list(
query.nodecontents(
ll.getElementsByTagName("ns1:data"))):
un.append('cas', sy, self.parent.confid,
self.parent.sourceid)
# pdb.set_trace()
# formula
if searchResults.getElementsByTagName('ns1:Formulae'):
form = searchResults.getElementsByTagName('ns1:Formulae')[0]
for sy in list(
query.nodecontents(form.getElementsByTagName("ns1:data"))):
un.append('formula', sy, self.parent.mm.confidence['weak'], \
self.parent.sourceid)
return (un)
def pubchem2mask(self, docSum):
""" turn a docsum node in to a mask
"""
un = mask({}, self.parent.mm.idpatterns)
cid = next(mmquery.nodecontents(docSum.getElementsByTagName("Id")))
un.append('cid', cid, self.parent.confid, self.parent.sourceid)
p = re.compile('(<a href[^>]*>)|(</a>)|(ligand)|(,)')
cnf = self.parent.confid
src = self.parent.sourceid
weak = self.parent.mm.confidence['weak']
for item in docSum.getElementsByTagName('Item'):
if len(item.childNodes) == 0:
continue
nVal = item.childNodes[0].nodeValue
if item.getAttribute('Name') == 'SynonymList':
synonyms = list(mmquery.nodecontents(item.childNodes))
for s in synonyms:
s = p.sub('', s.strip())
if 'kegg' in guessTable(s, self.parent.mm.idpatterns):
un.append('kegg', s, cnf, src)
continue
if 'cas' in guessTable(s, self.parent.mm.idpatterns):
un.append('cas', s, cnf, src)
continue
if 'chebi' in guessTable(s, self.parent.mm.idpatterns):
un.append('chebi', s, cnf, src)
continue
if 'inchi' in guessTable(s, self.parent.mm.idpatterns):
un.append('inchi', s, cnf, src)
continue
else:
un.append('synonym', s, weak, src)
elif item.getAttribute('Name') == 'IUPACName':
un.append('iupac', nVal, cnf, src)
elif item.getAttribute('Name') == 'CanonicalSmile':
un.append('smiles', nVal, cnf, src)
elif item.getAttribute('Name') == 'CanonicalSmile':
un.append('smiles', nVal, cnf, src)
elif item.getAttribute('Name') == 'InChIKey':
un.append('inchikey', nVal, cnf, src)
# annotation section
elif item.getAttribute('Name') == 'MolecularFormula':
un.append('formula', nVal, weak, src)
elif item.getAttribute('Name') == 'MolecularWeight':
un.append('weight', nVal, weak, src)
elif item.getAttribute('Name') == 'TotalFormalCharge':
un.append('totalcharge', nVal, weak, src)
elif item.getAttribute('Name') == 'XLogP':
un.append('xlogp', nVal, weak, src)
elif item.getAttribute('Name') == 'XLogP':
un.append('xlogp', nVal, weak, src)
elif item.getAttribute('Name') == 'HydrogenBondDonorCount':
un.append('hbonddonor', nVal, weak, src)
elif item.getAttribute('Name') == 'HydrogenBondAcceptorCount':
un.append('hbondacceptor', nVal, weak, src)
elif item.getAttribute('Name') == 'HeavyAtomCount':
un.append('heavyatom', nVal, weak, src)
elif item.getAttribute('Name') == 'TPSA':
un.append('tpsa', nVal, weak, src)
return (un)
def chebi2mask(self, mm, chebiId):
"""
get a mask containing the info associated with a chebi id
2. use getComplete to fetch the contents of the relevant entries
"""
ba = chebiId
un = mask({}, mm.idpatterns)
qUrl = self.getComplete + "chebiId=" + str(chebiId)
qRes = self.parent.getUrl(qUrl)
if not qRes:
return (mask({}))
searchResults = xml.dom.minidom.parse(qRes)
if not searchResults:
return (mask({}))
if searchResults.getElementsByTagName('ns1:return'):
retList = searchResults.getElementsByTagName('ns1:return')[0]
# found non-existent chebiId
else:
# this would have been reasonable but chebi has problems so that some
# entries exist, but cant be fetched. valid chebi but we cant query
# for it,
# hence, skip completely
# delmask = mask({})
# delmask.append('chebi', chebiId)
# mm.brandish(delmask)
return (mask({}))
# chebiid
if retList.getElementsByTagName("ns1:chebiId"):
chebiId = list(query.nodecontents(retList.getElementsByTagName("ns1:chebiId")))[0]
un.append('chebi', chebiId.replace("CHEBI:", ""),
self.parent.confid, self.parent.sourceid)
# smiles
if retList.getElementsByTagName("ns1:smiles"):
un.append('smiles', list(query.nodecontents(retList.getElementsByTagName("ns1:smiles")))[0],
self.parent.confid, self.parent.sourceid)
# synonym
if searchResults.getElementsByTagName('ns1:Synonyms'):
syns = searchResults.getElementsByTagName('ns1:Synonyms')[0]
for sy in list(query.nodecontents(syns.getElementsByTagName("ns1:data"))):
un.append('synonym', sy, self.parent.mm.confidence['weak'], \
self.parent.sourceid)
# inchi
if retList.getElementsByTagName("ns1:inchi"):
un.append('inchi', list(query.nodecontents(retList.getElementsByTagName("ns1:inchi")))[0],
self.parent.confid, self.parent.sourceid)
# iupac
if searchResults.getElementsByTagName('ns1:IupacNames'):
syns = searchResults.getElementsByTagName('ns1:IupacNames')[0]
for sy in list(query.nodecontents(syns.getElementsByTagName("ns1:data"))):
un.append('iupac', sy, self.parent.confid, \
self.parent.sourceid)
# kegg
for ll in searchResults.getElementsByTagName('ns1:DatabaseLinks'):
if list(query.nodecontents(ll.getElementsByTagName("ns1:type")))[0] == 'KEGG COMPOUND accession':
for sy in list(query.nodecontents(ll.getElementsByTagName("ns1:data"))):
un.append('kegg', sy, self.parent.confid, self.parent.sourceid)
# cas
for ll in searchResults.getElementsByTagName('ns1:RegistryNumbers'):
if list(query.nodecontents(ll.getElementsByTagName("ns1:type")))[0] == 'CAS Registry Number':
for sy in list(query.nodecontents(ll.getElementsByTagName("ns1:data"))):
un.append('cas', sy, self.parent.confid, self.parent.sourceid)
# pdb.set_trace()
# formula
if searchResults.getElementsByTagName('ns1:Formulae'):
form = searchResults.getElementsByTagName('ns1:Formulae')[0]
for sy in list(query.nodecontents(form.getElementsByTagName("ns1:data"))):
un.append('formula', sy, self.parent.mm.confidence['weak'], \
self.parent.sourceid)
return (un)
