def toMolFile(self,
identifier,
identifierType,
molfilePath=None,
fmt="mol",
**kwargs):
"""Create molfile (fmt) from InChI, SMILES descriptors or PDB identifier."""
try:
molfilePath = molfilePath if molfilePath else self.__makeMolfilePath(
fmt=fmt)
oeio = OeIoUtils()
if identifierType.lower() in ["smiles"]:
oeMol = oeio.smilesToMol(identifier)
oeMol.SetTitle("From SMILES")
elif identifierType.lower() in ["inchi"]:
oeMol = oeio.inchiToMol(identifier)
oeMol.SetTitle("From InChI")
elif identifierType.lower() in ["identifierpdb"]:
ccsw = ChemCompSearchWrapper()
oesmP = ccsw.getSearchMoleculeProvider()
oeMol = oesmP.getMol(identifier)
#
ok = self.__toMolFile(oeMol, molfilePath, **kwargs)
return molfilePath if ok else None
except Exception as e:
logger.exception("Failing with %s", str(e))
return None
def alignMoleculePair(self,
refIdentifier,
refIdentifierType,
fitIdentifier,
fitIdentifierType,
imagePath=None,
**kwargs):
"""Create aligned depiction for a target molecule InChI, SMILES descriptors or PDB identifier."""
try:
imagePath = imagePath if imagePath else self.__makeImagePath()
oeio = OeIoUtils()
ccsw = ChemCompSearchWrapper()
oesmP = ccsw.getSearchMoleculeProvider()
# ---
if refIdentifierType.lower() in ["smiles"]:
oeMolRef = oeio.smilesToMol(refIdentifier)
elif refIdentifierType.lower() in ["inchi"]:
oeMolRef = oeio.inchiToMol(refIdentifier)
elif refIdentifierType.lower() in ["identifierpdb"]:
oeMolRef = oesmP.getMol(refIdentifier)
#
if fitIdentifierType.lower() in ["smiles"]:
oeMolFit = oeio.smilesToMol(fitIdentifier)
elif fitIdentifierType.lower() in ["inchi"]:
oeMolFit = oeio.inchiToMol(fitIdentifier)
elif fitIdentifierType.lower() in ["identifierpdb"]:
oeMolFit = oesmP.getMol(fitIdentifier)
# ---
logger.info("oeMolRef atoms %r", oeMolRef.NumAtoms())
logger.info("oeMolFit atoms %r", oeMolFit.NumAtoms())
displayIdRef = "Ref"
displayIdFit = "Fit"
ok = self.__depictAlignedPair(oeMolRef, displayIdRef, oeMolFit,
displayIdFit, imagePath, **kwargs)
return imagePath if ok else None
except Exception as e:
logger.exception("Failing with %s", str(e))
return None
def testDepictSMILES(self):
"""Test case - create depiction from SMILES descriptor."""
try:
imagePath = os.path.join(self.__workPath, "benzene-from-smi.svg")
oeio = OeIoUtils()
oeMol = oeio.smilesToMol("c1ccccc1")
oed = OeDepict()
oed.setMolTitleList([("benzene", oeMol, "Title for benzene")])
oed.setDisplayOptions(labelAtomName=False,
labelAtomCIPStereo=True,
labelBondCIPStereo=True,
labelAtomIndex=False,
labelBondIndex=False,
bondDisplayWidth=1.0)
oed.setGridOptions(rows=1, cols=1)
oed.prepare()
oed.write(imagePath)
except Exception as e:
logger.exception("Failing with %s", str(e))
self.fail()
def depictMolecule(self,
identifier,
identifierType,
imagePath=None,
**kwargs):
"""Create depiction from InChI, SMILES descriptors or PDB identifier."""
try:
imagePath = imagePath if imagePath else self.__makeImagePath()
oeio = OeIoUtils()
if identifierType.lower() in ["smiles"]:
oeMol = oeio.smilesToMol(identifier)
elif identifierType.lower() in ["inchi"]:
oeMol = oeio.inchiToMol(identifier)
elif identifierType.lower() in ["identifierpdb"]:
ccsw = ChemCompSearchWrapper()
oesmP = ccsw.getSearchMoleculeProvider()
oeMol = oesmP.getMol(identifier)
#
ok = self.__depictOne(oeMol, imagePath, **kwargs)
return imagePath if ok else None
except Exception as e:
logger.exception("Failing with %s", str(e))
return None
def testSssWithFingerPrintFromDescriptor(self):
oemp = OeMoleculeProvider(**self.__myKwargs)
ok = oemp.testCache()
ccmP = ChemCompIndexProvider(**self.__myKwargs)
ccIdxD = ccmP.getIndex()
ok = ccmP.testCache(minCount=self.__minCount)
self.assertTrue(ok)
limitPerceptions = False
# minFpScore = 0.5
maxFpResults = 50
matchOpts = "graph-relaxed"
numMols = 20
oeioU = OeIoUtils()
oesU = OeSearchUtils(oemp, fpTypeList=self.__fpTypeList)
missTupL = []
missedD = {}
missedFpD = {}
# ----
startTime = time.time()
for ccId, ccD in list(ccIdxD.items())[:numMols]:
for buildType in [
"oe-iso-smiles", "oe-smiles", "acdlabs-smiles",
"cactvs-iso-smiles", "cactvs-smiles", "inchi"
]:
if buildType in ccD:
logger.debug("Search %s %r", ccId, ccD[buildType])
if buildType in ["inchi"]:
oemf = OeMoleculeFactory()
oemf.setDescriptor(ccD["inchi"], "inchi", ccId)
ok = oemf.build(molBuildType="inchi",
limitPerceptions=limitPerceptions)
if not ok:
logger.info("%s build failed with InChI %r", ccId,
ccD["inchi"])
else:
oeMol = oemf.getMol()
if oemf.getInChI() != ccD["inchi"]:
logger.info(
"%s regenerated InChI differs\n%r\n%s",
ccId, ccD["inchi"], oemf.getInChI())
else:
oeMol = oeioU.smilesToMol(
ccD[buildType], limitPerceptions=limitPerceptions)
if not oeMol:
continue
maxHits = 0
minHits = maxFpResults
selfHit = False
for fpType, minFpScore in self.__fpTypeCuttoffList:
retStatus, mL = oesU.searchSubStructureWithFingerPrint(
oeMol,
fpType,
minFpScore,
maxFpResults,
matchOpts=matchOpts)
self.assertTrue(retStatus)
logger.debug("%s fpType %r hits %d", ccId, fpType,
len(mL))
maxHits = max(maxHits, len(mL))
minHits = min(minHits, len(mL))
matchedSelf = self.__resultContains(ccId, mL)
selfHit = selfHit or matchedSelf
if not matchedSelf:
missedFpD.setdefault(ccId, []).append(
(buildType, fpType, len(mL)))
if not selfHit:
missedD.setdefault(ccId, []).append(buildType)
logger.info("%s (%r) buildType %r min hits %d max hits %d",
ccId, selfHit, buildType, minHits, maxHits)
else:
logger.info("%s missing descriptor %r", ccId, buildType)
#
for ccId, missL in missedD.items():
logger.info("%s missed list %r", ccId, missL)
if ccId in missedFpD:
logger.info("%s unmatched for fpTypes %r", ccId,
missedFpD[ccId])
# ----
doDepict = False
if doDepict:
mD = {}
for missTup in missTupL:
mD.setdefault(missTup[0], []).append(missTup[1])
for ccId, buildTypeL in mD.items():
idxD = ccIdxD[ccId]
if "oe-iso-smiles" in idxD:
for buildType in buildTypeL:
self.__displayAlignedDescriptorPair(
ccId,
idxD["oe-iso-smiles"],
"oe-iso-smiles",
idxD[buildType],
buildType,
title=None,
limitPerceptions=True)
logger.info("%s fingerprints search on %d in (%.4f seconds)",
len(self.__fpTypeList), numMols,
time.time() - startTime)