def searchSubStructureWithFingerPrint(self, oeQueryMol, fpType, minFpScore, maxFpResults, matchOpts="graph-relaxed"):
"""Return graph match search results for the input OE molecule using fingerprint pre-filtering.
Args:
oeQueryMol (OEmol): OE graph molecule
fpTypeCutoffList (list): [(finger print type, min score),...]
maxFpResults (int): maximum number of finger print results returned
matchOpts (str, optional): graph match criteria type (graph-strict|graph-relaxed). Defaults to "graph-relaxed".
Returns:
(bool, list, list): status, graph match and finger match lists of type (MatchResults)
"""
hL = []
retStatus = True
try:
fpDb = self.__fpDbD[fpType]
opts = oegraphsim.OEFPDatabaseOptions(maxFpResults, oegraphsim.OESimMeasure_Tanimoto)
if minFpScore:
opts.SetCutoff(minFpScore)
scores = fpDb.GetSortedScores(oeQueryMol, opts)
idxList = [si.GetIdx() for si in scores]
retStatus, hL = self.searchSubStructure(oeQueryMol, idxList=idxList, reverseFlag=False, matchOpts=matchOpts)
except Exception as e:
retStatus = False
logger.exception("Failing with %s", str(e))
return retStatus, hL
def getFingerPrintScores(self, oeQueryMol, fpType, minFpScore, maxFpResults):
"""Return finger print search scores for the input OE molecule.
Args:
oeQueryMol (OEmol): OE graph molecule
fpType (str): fingerprint type [TREE,PATH,MACCS,CIRCULAR,LINGO]
fpMinScore (float): min fingerprint match score (0.0-1.0)
maxFpResults (int): maximum number of finger print results returned
Returns:
(bool, list): status, finger match lists of type (MatchResults)
"""
hL = []
retStatus = True
try:
fpDb = self.__fpDbD[fpType]
opts = oegraphsim.OEFPDatabaseOptions(maxFpResults, oegraphsim.OESimMeasure_Tanimoto)
if minFpScore:
opts.SetCutoff(minFpScore)
scores = fpDb.GetSortedScores(oeQueryMol, opts)
hL = [MatchResults(ccId=self.__oeMolDb.GetTitle(si.GetIdx()), searchType="fp", fpType=fpType, fpScore=si.GetScore(), oeIdx=si.GetIdx()) for si in scores]
except Exception as e:
retStatus = False
logger.exception("Failing with %s", str(e))
return retStatus, hL
def searchFingerPrints(self, oeQueryMol, fpType, minFpScore=None, maxFpResults=50, annotateMols=False, verbose=False):
hL = []
retStatus = True
try:
fpDb = self.__fpDbD[fpType] if fpType in self.__fpDbD else {}
if not fpDb:
retStatus = False
return retStatus, hL
#
opts = oegraphsim.OEFPDatabaseOptions(maxFpResults, oegraphsim.OESimMeasure_Tanimoto)
if minFpScore:
opts.SetCutoff(minFpScore)
#
if verbose:
logger.info("Using %d fingerprint %s type %s", fpDb.NumFingerPrints(), fpType, fpDb.GetFPTypeBase().GetFPTypeString())
startTime = time.time()
#
scores = fpDb.GetSortedScores(oeQueryMol, opts)
oeMol = oechem.OEGraphMol()
for si in scores:
if self.__oeMolDb.GetMolecule(oeMol, si.GetIdx()):
ccId = self.__oeMolDb.GetTitle(si.GetIdx())
if annotateMols:
tS = "For %s index %r %r similarity score %.4f " % (ccId, si.GetIdx(), self.__idxTitleD[si.GetIdx()], si.GetScore())
oechem.OESetSDData(oeMol, fpType, tS)
hL.append(MatchResults(ccId=ccId, oeMol=oeMol, searchType="fp", fpType=fpType, fpScore=si.GetScore()))
if verbose:
endTime = time.time()
logger.info("Fingerprint %s returning %d hits (%.4f sec)", fpType, len(hL), endTime - startTime)
except Exception as e:
retStatus = False
logger.exception("Failing fpType %r with %s", fpType, str(e))
return retStatus, hL
def main(argv=[__name__]):
itf = oechem.OEInterface()
oechem.OEConfigure(itf, InterfaceData)
defopts = oegraphsim.OEFPDatabaseOptions(10,
oegraphsim.OESimMeasure_Tanimoto)
oegraphsim.OEConfigureFPDatabaseOptions(itf, defopts)
oegraphsim.OEConfigureFPDatabaseMemoryType(itf)
if not oechem.OEParseCommandLine(itf, argv):
return 0
qfname = itf.GetString("-query")
mfname = itf.GetString("-molfname")
ffname = itf.GetString("-fpdbfname")
ofname = itf.GetString("-out")
# initialize databases
timer = oechem.OEWallTimer()
timer.Start()
ifs = oechem.oemolistream()
if not ifs.open(qfname):
oechem.OEThrow.Fatal("Cannot open input file!")
query = oechem.OEGraphMol()
if not oechem.OEReadMolecule(ifs, query):
oechem.OEThrow.Fatal("Cannot read query molecule!")
moldb = oechem.OEMolDatabase()
if not moldb.Open(mfname):
oechem.OEThrow.Fatal("Cannot open molecule database!")
memtype = oegraphsim.OEGetFPDatabaseMemoryType(itf)
fpdb = oegraphsim.OEFastFPDatabase(ffname, memtype)
if not fpdb.IsValid():
oechem.OEThrow.Fatal("Cannot open fingerprint database!")
nrfps = fpdb.NumFingerPrints()
memtypestr = fpdb.GetMemoryTypeString()
ofs = oechem.oemolostream()
if not ofs.open(ofname):
oechem.OEThrow.Fatal("Cannot open output file!")
if not oegraphsim.OEAreCompatibleDatabases(moldb, fpdb):
oechem.OEThrow.Fatal("Databases are not compatible!")
oechem.OEThrow.Info("%5.2f sec to initialize databases" % timer.Elapsed())
fptype = fpdb.GetFPTypeBase()
oechem.OEThrow.Info("Using fingerprint type %s" % fptype.GetFPTypeString())
opts = oegraphsim.OEFPDatabaseOptions()
oegraphsim.OESetupFPDatabaseOptions(opts, itf)
# search fingerprint database
timer.Start()
scores = fpdb.GetSortedScores(query, opts)
oechem.OEThrow.Info("%5.2f sec to search %d fingerprints %s" %
(timer.Elapsed(), nrfps, memtypestr))
timer.Start()
nrhits = 0
hit = oechem.OEGraphMol()
for si in scores:
if moldb.GetMolecule(hit, si.GetIdx()):
nrhits += 1
oechem.OESetSDData(hit, "Similarity score", "%.2f" % si.GetScore())
oechem.OEWriteMolecule(ofs, hit)
oechem.OEThrow.Info("%5.2f sec to write %d hits" %
(timer.Elapsed(), nrhits))
return 0
def main(argv=[__name__]):
itf = oechem.OEInterface()
oechem.OEConfigure(itf, InterfaceData)
defopts = oegraphsim.OEFPDatabaseOptions(10,
oegraphsim.OESimMeasure_Tanimoto)
oegraphsim.OEConfigureFPDatabaseOptions(itf, defopts)
oegraphsim.OEConfigureFingerPrint(
itf, oegraphsim.OEGetFPType(oegraphsim.OEFPType_Tree))
if not oechem.OEParseCommandLine(itf, argv):
return 0
qfname = itf.GetString("-query")
mfname = itf.GetString("-molfname")
ofname = itf.GetString("-out")
# initialize databases
timer = oechem.OEWallTimer()
timer.Start()
ifs = oechem.oemolistream()
if not ifs.open(qfname):
oechem.OEThrow.Fatal("Cannot open input file!")
query = oechem.OEGraphMol()
if not oechem.OEReadMolecule(ifs, query):
oechem.OEThrow.Fatal("Cannot read query molecule!")
moldb = oechem.OEMolDatabase()
if not moldb.Open(mfname):
oechem.OEThrow.Fatal("Cannot open molecule database!")
ofs = oechem.oemolostream()
if not ofs.open(ofname):
oechem.OEThrow.Fatal("Cannot open output file!")
fptype = oegraphsim.OESetupFingerPrint(itf)
oechem.OEThrow.Info("Using fingerprint type %s" % fptype.GetFPTypeString())
fpdb = oegraphsim.OEFPDatabase(fptype)
emptyfp = oegraphsim.OEFingerPrint()
emptyfp.SetFPTypeBase(fptype)
nrmols = moldb.GetMaxMolIdx()
mol = oechem.OEGraphMol()
for idx in range(0, nrmols):
if moldb.GetMolecule(mol, idx):
fpdb.AddFP(mol)
else:
fpdb.AddFP(emptyfp)
nrfps = fpdb.NumFingerPrints()
oechem.OEThrow.Info("%5.2f sec to initialize databases" % timer.Elapsed())
opts = oegraphsim.OEFPDatabaseOptions()
oegraphsim.OESetupFPDatabaseOptions(opts, itf)
# search fingerprint database
timer.Start()
scores = fpdb.GetSortedScores(query, opts)
oechem.OEThrow.Info("%5.2f sec to search %d fingerprints" %
(timer.Elapsed(), nrfps))
timer.Start()
hit = oechem.OEGraphMol()
for si in scores:
if moldb.GetMolecule(hit, si.GetIdx()):
oechem.OEWriteMolecule(ofs, hit)
oechem.OEThrow.Info("%5.2f sec to write %d hits" %
(timer.Elapsed(), opts.GetLimit()))
return 0