def SerializeInput(itf):
ims = oechem.oemolistream()
if not ims.open(itf.GetString("-input")):
oechem.OEThrow.Fatal("Unable to open %s for reading:" +
itf.GetString("-input"))
oechem.OEThrow.SetLevel(oechem.OEErrorLevel_Warning)
# @ <SNIPPET-MATCHEDPAIRANALYZER-EXPORT-COMPRESS-MMP>
# create analyzer class with defaults
# - compression option disabled by default
mmpAnalyzer = oemedchem.OEMatchedPairAnalyzer()
# for serialization, enable export compression to
# remove singleton index nodes by modifying analyzer
mmpAnalyzer.ModifyOptions(oemedchem.OEMatchedPairOptions_ExportCompression,
0)
# @ </SNIPPET-MATCHEDPAIRANALYZER-EXPORT-COMPRESS-MMP>
# @ <SNIPPET-MATCHEDPAIRANALYZER-EXPORT-COMPRESS-OPTS>
# create options class with defaults
# - compression option disabled by default
mmpOpts = oemedchem.OEMatchedPairAnalyzerOptions()
# for serialization, enable export compression to
# remove singleton index nodes by modifying analyzer
mmpOpts.SetOptions(oemedchem.OEMatchedPairOptions_Default
| oemedchem.OEMatchedPairOptions_ExportCompression)
# create analyzer class with compression option enabled
mmpAnalyzer = oemedchem.OEMatchedPairAnalyzer(mmpOpts)
# @ </SNIPPET-MATCHEDPAIRANALYZER-EXPORT-COMPRESS-OPTS>
# @ <SNIPPET-MATCHEDPAIRANALYZER-SERIALIZE-INDEX>
mmp = oemedchem.OEMatchedPairAnalyzer()
for recindex, mol in enumerate(ims.GetOEGraphMols()):
status = mmp.AddMol(mol, recindex)
if status != recindex:
oechem.OEThrow.Warning(
"{0}: molecule indexing error, status={1}".format(
recindex, oemedchem.OEMatchedPairIndexStatusName(status)))
print("Index complete, matched pairs = {0}".format(mmp.NumMatchedPairs()))
# check for output filename with .mmpidx extension
mmpexport = itf.GetString("-output")
if not oemedchem.OEIsMatchedPairAnalyzerFileType(mmpexport):
oechem.OEThrow.Info('Not a valid matched pair index output file, ' +
mmpexport)
elif not oemedchem.OEWriteMatchedPairAnalyzer(mmpexport, mmp):
oechem.OEThrow.Fatal("Index serialization failed")
else:
oechem.OEThrow.Info("Index serialization complete")
# @ </SNIPPET-MATCHEDPAIRANALYZER-SERIALIZE-INDEX>
print("Index serialization complete")
mmpimport = mmpexport
# now try to reload serialized index
# @ <SNIPPET-MATCHEDPAIRANALYZER-DESERIALIZE-INDEX>
mmp = oemedchem.OEMatchedPairAnalyzer()
if not oemedchem.OEIsMatchedPairAnalyzerFileType(mmpimport):
oechem.OEThrow.Fatal('Not a valid matched pair index input file, ' +
mmpimport)
elif not oemedchem.OEReadMatchedPairAnalyzer(mmpimport, mmp):
oechem.OEThrow.Fatal("Index deserialization failed")
else:
oechem.OEThrow.Info("Index deserialization complete")
# @ </SNIPPET-MATCHEDPAIRANALYZER-DESERIALIZE-INDEX>
print("Index deserialization complete")
return True
def FindSimpleMatchedPairs(itf):
ims = oechem.oemolistream()
if not ims.open(itf.GetString("-input")):
oechem.OEThrow.Fatal("Unable to open %s for reading: " +
itf.GetString("-input"))
maxrecs = itf.GetInt("-maxrec")
oechem.OEThrow.SetLevel(oechem.OEErrorLevel_Warning)
# @ <SNIPPET-FINDSIMPLEMATCHEDPAIRS-EXAMPLE>
# create options class with defaults
mmpOpts = oemedchem.OEMatchedPairAnalyzerOptions()
# for 'simple' pairs, alter default indexing options
# - single cuts only, heavy atom substituents only (HMember indexing off)
mmpOpts.SetOptions(oemedchem.OEMatchedPairOptions_SingleCuts
| oemedchem.OEMatchedPairOptions_ComboCuts
| oemedchem.OEMatchedPairOptions_UniquesOnly)
# - limit substituent size to no more than 20% of input structure
mmpOpts.SetIndexableFragmentRange(80., 100.)
# create analyzer class with nondefault options
mmpAnalyzer = oemedchem.OEMatchedPairAnalyzer(mmpOpts)
# ignore common index status returns
sIgnoreStatus = 'FragmentRangeFilter,DuplicateStructure,'
sIgnoreStatus += 'FragmentationLimitFilter,HeavyAtomFilter'
# index the input structures
for recindex, mol in enumerate(ims.GetOEGraphMols(), start=1):
# consider only the largest input fragment
oechem.OEDeleteEverythingExceptTheFirstLargestComponent(mol)
# ignore stereochemistry
oechem.OEUncolorMol(mol, (oechem.OEUncolorStrategy_RemoveAtomStereo
| oechem.OEUncolorStrategy_RemoveBondStereo))
# explicitly provide a 1-based index to refer to indexed structures
# - to allow references back to external data elsewhere
status = mmpAnalyzer.AddMol(mol, recindex)
if status != recindex:
if not oemedchem.OEMatchedPairIndexStatusName(
status) in sIgnoreStatus:
oechem.OEThrow.Warning(
"{0}: molecule indexing error, status={1}".format(
recindex,
oemedchem.OEMatchedPairIndexStatusName(status)))
# if limiting input, quit after limit
if maxrecs and recindex >= maxrecs:
break
print("Index complete, matched pairs = {0}".format(
mmpAnalyzer.NumMatchedPairs()))
# specify how transforms are extracted (direction and allowed properties)
extractMode = (
oemedchem.OEMatchedPairTransformExtractMode_Sorted
| oemedchem.OEMatchedPairTransformExtractMode_NoSMARTS
| oemedchem.OEMatchedPairTransformExtractMode_AddMCSCorrespondence)
extractOptions = oemedchem.OEMatchedPairTransformExtractOptions()
# specify amount of chemical context at the site of the substituent change
# in the transform
extractOptions.SetContext(oemedchem.OEMatchedPairContext_Bond0)
extractOptions.SetOptions(extractMode)
# walk the transforms and print the matched pairs
xfmidx = 0
for mmpxform in oemedchem.OEMatchedPairGetTransforms(
mmpAnalyzer, extractOptions):
xfmidx += 1
print("{0:2} {1}".format(xfmidx, mmpxform.GetTransform()))
# dump matched molecular pairs and index identifiers
# (recindex from indexing loop above)
for mmppair in mmpxform.GetMatchedPairs():
print("\tmatched pair molecule indices=({0},{1})".format(
mmppair.FromIndex(), mmppair.ToIndex()))
# @ </SNIPPET-FINDSIMPLEMATCHEDPAIRS-EXAMPLE>
return True
def MMPTransformList(itf):
# check MMP index
mmpimport = itf.GetString("-mmpindex")
if not oemedchem.OEIsMatchedPairAnalyzerFileType(mmpimport):
oechem.OEThrow.Fatal(
'Not a valid matched pair index input file, {}'.format(mmpimport))
# load MMP index
mmp = oemedchem.OEMatchedPairAnalyzer()
if not oemedchem.OEReadMatchedPairAnalyzer(mmpimport, mmp):
oechem.OEThrow.Fatal("Unable to load index {}".format(mmpimport))
if not mmp.NumMols():
oechem.OEThrow.Fatal(
'No records in loaded MMP index file: {}'.format(mmpimport))
if not mmp.NumMatchedPairs():
oechem.OEThrow.Fatal(
'No matched pairs found in MMP index file, ' +
'use -fragGe,-fragLe options to extend indexing range')
# request a specific context for the transform activity, here 0-bonds
chemctxt = oemedchem.OEMatchedPairContext_Bond0
askcontext = itf.GetString("-context")[:1]
if askcontext == '0':
chemctxt = oemedchem.OEMatchedPairContext_Bond0
elif askcontext == '1':
chemctxt = oemedchem.OEMatchedPairContext_Bond1
elif askcontext == '2':
chemctxt = oemedchem.OEMatchedPairContext_Bond2
elif askcontext == '3':
chemctxt = oemedchem.OEMatchedPairContext_Bond3
elif askcontext == 'a' or askcontext == 'A':
chemctxt = oemedchem.OEMatchedPairContext_AllBonds
else:
oechem.OEThrow.Fatal("Invalid context specified: " + askcontext +
", only 0|1|2|3|A allowed")
bPrintTransforms = itf.GetBool("-printlist")
# if a data field was specified, retrieve the SD data field name
field = None
if itf.HasString("-datafield"):
field = itf.GetString("-datafield")
if not bPrintTransforms and field is None:
oechem.OEThrow.Info(
'Specify one of -datafield or -printlist, otherwise nothing to do!'
)
return
extractOptions = oemedchem.OEMatchedPairTransformExtractOptions()
# specify amount of chemical context at the site of the substituent change
extractOptions.SetContext(chemctxt)
# controls how transforms are extracted (direction and allowed properties)
extractOptions.SetOptions(
oemedchem.OEMatchedPairTransformExtractMode_Sorted
| oemedchem.OEMatchedPairTransformExtractMode_NoSMARTS)
# walk the transforms from the indexed matched pairs
xforms = []
xfmidx = 0
for mmpxform in oemedchem.OEMatchedPairGetTransforms(mmp, extractOptions):
xfmidx += 1
if bPrintTransforms:
print("{0:2} {1}".format(xfmidx, mmpxform.GetTransform()))
# compute delta property
mmpidx = 0
prop = []
for mmppair in mmpxform.GetMatchedPairs():
mmpidx += 1
mmpinfo = "\t{0:2}: ({1:2},{2:2})".format(mmpidx,
mmppair.FromIndex(),
mmppair.ToIndex())
for tag in mmppair.GetDataTags():
mmpinfo = mmpinfo + " {0}=({1},{2})".format(
tag, mmppair.GetFromSDData(tag), mmppair.GetToSDData(tag))
if tag == field:
fromData = None
toData = None
try:
fromData = float(mmppair.GetFromSDData(tag))
except ValueError:
fromData = None
try:
toData = float(mmppair.GetToSDData(tag))
except ValueError:
toData = None
if fromData is not None and toData is not None:
prop.append(toData - fromData)
if bPrintTransforms:
print(mmpinfo)
# skip if property not found
if len(prop):
xforms.append(
MMPXform(mmpxform, average(prop), stddev(prop), len(prop)))
if not field:
return 0
if field and not len(xforms):
oechem.OEThrow.Error(
"No matched pairs found with {} data".format(field))
print("")
print("*** Transforms sorted by delta ({})".format(field))
xforms.sort(key=lambda c: -abs(float(c.avg)))
idx = 0
for xfm in xforms:
idx += 1
if (extractOptions.GetOptions()
& oemedchem.OEMatchedPairTransformExtractMode_NoSMARTS) != 0:
# not 'invertable' if SMARTS qualifiers were applied
if xfm.avg < 0.:
xfm.avg = -1. * xfm.avg
xfm.xform.Invert()
print("{0:2} {2}=(avg={3:.2f},stdev={4:.2f},num={5}) {1}".format(
idx, xfm.xform.GetTransform(), field, xfm.avg, xfm.std, xfm.num))
def MMPIndex(itf):
# output index file
mmpindexfile = itf.GetString("-output")
if not oemedchem.OEIsMatchedPairAnalyzerFileType(mmpindexfile):
oechem.OEThrow.Fatal("Output file is not a matched pair index type - \
needs .mmpidx extension: {}"
.format(mmpindexfile))
# create options class with defaults
mmpopts = oemedchem.OEMatchedPairAnalyzerOptions()
# set up options from command line
if not oemedchem.OESetupMatchedPairIndexOptions(mmpopts, itf):
oechem.OEThrow.Fatal("Error setting matched pair indexing options!")
# input structures to index
ifsindex = oechem.oemolistream()
if not ifsindex.open(itf.GetString("-input")):
oechem.OEThrow.Fatal("Unable to open {} for reading"
.format(itf.GetString("-input")))
# get requested verbosity setting
verbose = itf.GetBool("-verbose")
vverbose = itf.GetBool("-vverbose")
if vverbose:
verbose = vverbose
maxrec = max(itf.GetInt("-maxrec"), 0)
statusrec = itf.GetInt("-status")
if itf.GetBool("-exportcompress"):
if not mmpopts.SetOptions(mmpopts.GetOptions() |
oemedchem.OEMatchedPairOptions_ExportCompression):
oechem.OEThrow.Warning("Error enabling export compression!")
stripstereo = itf.GetBool("-stripstereo")
stripsalts = itf.GetBool("-stripsalts")
keepFields = []
if itf.HasString("-keepSD"):
for field in itf.GetStringList("-keepSD"):
keepFields.append(field)
if verbose:
oechem.OEThrow.Info('Retaining SD data fields: {}'.format(' '.join(keepFields)))
alldata = itf.GetBool("-allSD")
cleardata = itf.GetBool("-clearSD")
if keepFields:
if verbose and (alldata or cleardata):
oechem.OEThrow.Info("Option -keepSD overriding -allSD, -clearSD")
alldata = False
cleardata = False
elif cleardata:
alldata = False
if verbose:
oechem.OEThrow.Info("Forced clearing of all input SD data")
elif alldata:
if verbose:
oechem.OEThrow.Info("Retaining all input SD data")
cleardata = False
elif verbose:
oechem.OEThrow.Info("No SD data handling option specified, -allSD assumed")
if cleardata:
keepFields = ['-CLEARSD']
elif alldata or not keepFields:
keepFields = ['-ALLSD']
if verbose:
if not mmpopts.HasIndexableFragmentHeavyAtomRange():
oechem.OEThrow.Info("Indexing all fragments")
else:
oechem.OEThrow.Info("Limiting fragment cores to {0:.2f}-{1:.2f}% of input molecules"
.format(mmpopts.GetIndexableFragmentRangeMin(),
mmpopts.GetIndexableFragmentRangeMax()))
if statusrec:
oechem.OEThrow.Info("Status output after every {0} records".format(statusrec))
if maxrec:
oechem.OEThrow.Info("Indexing a maximum of {0} records".format(maxrec))
if itf.GetBool("-exportcompress"):
oechem.OEThrow.Info("Removing singleton index nodes from index")
if stripstereo:
oechem.OEThrow.Info("Stripping stereo")
if stripsalts:
oechem.OEThrow.Info("Stripping salts")
if itf.GetBool("-clearSD"):
oechem.OEThrow.Info("Clearing all input SD data")
elif alldata:
oechem.OEThrow.Info("Retaining all input SD data")
elif keepFields:
oechem.OEThrow.Info('Retaining floating point SD data fields: {}'
.format(''.join(keepFields)))
# create indexing engine
mmp = oemedchem.OEMatchedPairAnalyzer(mmpopts)
# interpret SD fields as floating point data
validdata = FilterSDData(keepFields, True)
# add molecules to be indexed
record = 0
unindexed = 0
for mol in ifsindex.GetOEGraphMols():
if not alldata:
# filter the input molecule SD data based on allowed fields
validdata.FilterMolData(mol)
if stripsalts:
oechem.OEDeleteEverythingExceptTheFirstLargestComponent(mol)
if stripstereo:
oechem.OEUncolorMol(mol,
(oechem.OEUncolorStrategy_RemoveAtomStereo |
oechem.OEUncolorStrategy_RemoveBondStereo |
oechem.OEUncolorStrategy_RemoveGroupStereo))
status = mmp.AddMol(mol, record)
if status != record:
unindexed += 1
if vverbose:
oechem.OEThrow.Info('Input structure not added to index, record=%d status=%s' %
(record, oemedchem.OEMatchedPairIndexStatusName(status)))
record += 1
if maxrec and record >= maxrec:
break
if statusrec and (record % statusrec) == 0:
oechem.OEThrow.Info("Records: {} Indexed: {} Unindexed: {}"
.format(record, (record - unindexed), unindexed))
if not mmp.NumMols():
oechem.OEThrow.Fatal('No records in index structure file')
if not mmp.NumMatchedPairs():
oechem.OEThrow.Fatal('No matched pairs found from indexing, ' +
'use -fragGe,-fragLe options to extend indexing range')
if not oemedchem.OEWriteMatchedPairAnalyzer(mmpindexfile, mmp):
oechem.OEThrow.Fatal('Error serializing MMP index: {}'
.format(mmpindexfile))
# return some status information
oechem.OEThrow.Info("Records: {}, Indexed: {}, matched pairs: {:,d}"
.format(record,
mmp.NumMols(),
mmp.NumMatchedPairs()))
return 0
def MMPTransform(itf):
# input structure(s) to process
ifsmols = oechem.oemolistream()
if not ifsmols.open(itf.GetString("-input")):
oechem.OEThrow.Fatal("Unable to open %s for reading" %
itf.GetString("-input"))
# check MMP index
mmpimport = itf.GetString("-mmpindex")
if not oemedchem.OEIsMatchedPairAnalyzerFileType(mmpimport):
oechem.OEThrow.Fatal(
'Not a valid matched pair index input file, {}'.format(mmpimport))
# load MMP index
mmp = oemedchem.OEMatchedPairAnalyzer()
if not oemedchem.OEReadMatchedPairAnalyzer(mmpimport, mmp):
oechem.OEThrow.Fatal("Unable to load index {}".format(mmpimport))
if not mmp.NumMols():
oechem.OEThrow.Fatal(
'No records in loaded MMP index file: {}'.format(mmpimport))
if not mmp.NumMatchedPairs():
oechem.OEThrow.Fatal(
'No matched pairs found in MMP index file, ' +
'use -fragGe,-fragLe options to extend indexing range')
# output (transformed) structure(s)
ofs = oechem.oemolostream()
if not ofs.open(itf.GetString("-output")):
oechem.OEThrow.Fatal("Unable to open %s for writing" %
itf.GetString("-output"))
# request a specific context for the transform activity, here 0-bonds
chemctxt = oemedchem.OEMatchedPairContext_Bond0
askcontext = itf.GetString("-context")[:1]
if askcontext == '0':
chemctxt = oemedchem.OEMatchedPairContext_Bond0
elif askcontext == '1':
chemctxt = oemedchem.OEMatchedPairContext_Bond1
elif askcontext == '2':
chemctxt = oemedchem.OEMatchedPairContext_Bond2
elif askcontext == '3':
chemctxt = oemedchem.OEMatchedPairContext_Bond3
elif askcontext == 'a' or askcontext == 'A':
chemctxt = oemedchem.OEMatchedPairContext_AllBonds
else:
oechem.OEThrow.Fatal("Invalid context specified: " + askcontext +
", only 0|1|2|3|A allowed")
verbose = itf.GetBool("-verbose")
# return some status information
if verbose:
oechem.OEThrow.Info("{}: molecules: {:d}, matched pairs: {:,d}".format(
mmpimport, mmp.NumMols(), mmp.NumMatchedPairs()))
minpairs = itf.GetInt("-minpairs")
if minpairs > 1 and verbose:
oechem.OEThrow.Info(
'Requiring at least %d matched pairs to apply transformations' %
minpairs)
errs = None
if itf.GetBool("-nowarnings"):
errs = oechem.oeosstream()
oechem.OEThrow.SetOutputStream(errs)
orec = 0
ocnt = 0
for mol in ifsmols.GetOEGraphMols():
orec += 1
iter = oemedchem.OEMatchedPairApplyTransforms(mol, mmp, chemctxt,
minpairs)
if not iter.IsValid():
if verbose:
# as minpairs increases, fewer transformed mols are generated - output if requested
name = mol.GetTitle()
if not mol.GetTitle():
name = 'Record ' + str(orec)
oechem.OEThrow.Info("%s did not produce any output" % name)
continue
if errs is not None:
errs.clear()
for outmol in iter:
ocnt += 1
oechem.OEWriteMolecule(ofs, outmol)
if errs is not None:
errs.clear()
if not orec:
oechem.OEThrow.Fatal('No records in input structure file to transform')
if not ocnt:
oechem.OEThrow.Warning('No transformed structures generated')
print("Input molecules={} Output molecules={}".format(orec, ocnt))
return 0