def ScreenInDb(details, mol):
try:
probeFp = FingerprintMols.FingerprintMol(mol, **details.__dict__)
except Exception:
import traceback
FingerprintMols.error('Error: problems fingerprinting molecule.\n')
traceback.print_exc()
return []
if details.metric not in (DataStructs.TanimotoSimilarity, DataStructs.DiceSimilarity,
DataStructs.CosineSimilarity):
return ScreenFingerprints(details, data=GetFingerprints(details), mol=mol)
conn: DbConnect = _ConnectToDatabase(details)
if details.metric == DataStructs.TanimotoSimilarity:
func = 'rd_tanimoto'
elif details.metric == DataStructs.DiceSimilarity:
func = 'rd_dice'
elif details.metric == DataStructs.CosineSimilarity:
func = 'rd_cosine'
pkl = probeFp.ToBitString()
extraFields = f"{func}({DbModule.placeHolder},{details.fpColName}) as tani"
cmd = _ConstructSQL(details, extraFields=extraFields)
if details.doThreshold:
# we need to do a subquery here:
cmd = f"select * from ({cmd}) tmp where tani>{details.screenThresh}"
cmd += " order by tani desc"
if not details.doThreshold and details.topN > 0:
cmd += f" limit {details.topN}"
curs = conn.GetCursor()
curs.execute(cmd, (pkl, ))
return curs.fetchall()
def GetFingerprints(details):
""" returns an iterable sequence of fingerprints
each fingerprint will have a _fieldsFromDb member whose first entry is
the id.
"""
if details.dbName and details.tableName:
try:
conn = DbConnect(details.dbName, details.tableName)
if hasattr(details, 'dbUser'):
conn.user = details.dbUser
if hasattr(details, 'dbPassword'):
conn.password = details.dbPassword
except:
import traceback
FingerprintMols.error(
'Error: Problems establishing connection to database: %s|%s\n'
% (details.dbName, details.tableName))
traceback.print_exc()
cmd = _ConstructSQL(details, extraFields=details.fpColName)
curs = conn.GetCursor()
#curs.execute(cmd)
#print 'CURSOR:',curs,curs.closed
if _dataSeq:
suppl = _dataSeq(curs,
cmd,
depickle=not details.noPickle,
klass=DataStructs.ExplicitBitVect)
_dataSeq._conn = conn
else:
suppl = DbFpSupplier.ForwardDbFpSupplier(
data, fpColName=details.fpColName)
elif details.inFileName:
conn = None
try:
inF = open(details.inFileName, 'r')
except IOError:
import traceback
FingerprintMols.error('Error: Problems reading from file %s\n' %
(details.inFileName))
traceback.print_exc()
supple = []
done = 0
while not done:
try:
id, fp = cPickle.load(inF)
except:
done = 1
else:
fp._fieldsFromDb = [id]
suppl.append(fp)
else:
suppl = None
return suppl
def ScreenInDb(details, mol):
try:
probeFp = apply(FingerprintMols.FingerprintMol, (mol, ),
details.__dict__)
except:
import traceback
FingerprintMols.error('Error: problems fingerprinting molecule.\n')
traceback.print_exc()
return []
if details.dbName and details.tableName:
try:
conn = DbConnect(details.dbName, details.tableName)
if hasattr(details, 'dbUser'):
conn.user = details.dbUser
if hasattr(details, 'dbPassword'):
conn.password = details.dbPassword
except:
import traceback
FingerprintMols.error(
'Error: Problems establishing connection to database: %s|%s\n'
% (details.dbName, details.tableName))
traceback.print_exc()
if details.metric not in (DataStructs.TanimotoSimilarity,
DataStructs.DiceSimilarity,
DataStructs.CosineSimilarity):
data = GetFingerprints(details)
res = ScreenFingerprints(details, data, mol)
else:
res = []
if details.metric == DataStructs.TanimotoSimilarity:
func = 'rd_tanimoto'
pkl = probeFp.ToBitString()
elif details.metric == DataStructs.DiceSimilarity:
func = 'rd_dice'
pkl = probeFp.ToBitString()
elif details.metric == DataStructs.CosineSimilarity:
func = 'rd_cosine'
pkl = probeFp.ToBitString()
extraFields = "%s(%s,%s) as tani" % (func, DbModule.placeHolder,
details.fpColName)
cmd = _ConstructSQL(details, extraFields=extraFields)
if details.doThreshold:
# we need to do a subquery here:
cmd = "select * from (%s) tmp where tani>%f" % (
cmd, details.screenThresh)
cmd += " order by tani desc"
if not details.doThreshold and details.topN > 0:
cmd += " limit %d" % details.topN
curs = conn.GetCursor()
curs.execute(cmd, (pkl, ))
res = curs.fetchall()
return res
def GetFingerprints(details):
""" returns an iterable sequence of fingerprints
each fingerprint will have a _fieldsFromDb member whose first entry is
the id.
"""
if details.dbName and details.tableName:
try:
conn = DbConnect(details.dbName, details.tableName)
if hasattr(details, "dbUser"):
conn.user = details.dbUser
if hasattr(details, "dbPassword"):
conn.password = details.dbPassword
except Exception:
import traceback
FingerprintMols.error(
"Error: Problems establishing connection to database: %s|%s\n" % (details.dbName, details.tableName)
)
traceback.print_exc()
cmd = _ConstructSQL(details, extraFields=details.fpColName)
curs = conn.GetCursor()
# curs.execute(cmd)
# print 'CURSOR:',curs,curs.closed
if _dataSeq:
suppl = _dataSeq(curs, cmd, depickle=not details.noPickle, klass=DataStructs.ExplicitBitVect)
_dataSeq._conn = conn
else:
suppl = DbFpSupplier.ForwardDbFpSupplier(data, fpColName=details.fpColName)
elif details.inFileName:
conn = None
try:
inF = open(details.inFileName, "r")
except IOError:
import traceback
FingerprintMols.error("Error: Problems reading from file %s\n" % (details.inFileName))
traceback.print_exc()
suppl = []
done = 0
while not done:
try:
ID, fp = cPickle.load(inF)
except Exception:
done = 1
else:
fp._fieldsFromDb = [ID]
suppl.append(fp)
else:
suppl = None
return suppl
def ScreenInDb(details, mol):
try:
probeFp = apply(FingerprintMols.FingerprintMol, (mol,), details.__dict__)
except Exception:
import traceback
FingerprintMols.error("Error: problems fingerprinting molecule.\n")
traceback.print_exc()
return []
if details.dbName and details.tableName:
try:
conn = DbConnect(details.dbName, details.tableName)
if hasattr(details, "dbUser"):
conn.user = details.dbUser
if hasattr(details, "dbPassword"):
conn.password = details.dbPassword
except Exception:
import traceback
FingerprintMols.error(
"Error: Problems establishing connection to database: %s|%s\n" % (details.dbName, details.tableName)
)
traceback.print_exc()
if details.metric not in (DataStructs.TanimotoSimilarity, DataStructs.DiceSimilarity, DataStructs.CosineSimilarity):
data = GetFingerprints(details)
res = ScreenFingerprints(details, data, mol)
else:
res = []
if details.metric == DataStructs.TanimotoSimilarity:
func = "rd_tanimoto"
pkl = probeFp.ToBitString()
elif details.metric == DataStructs.DiceSimilarity:
func = "rd_dice"
pkl = probeFp.ToBitString()
elif details.metric == DataStructs.CosineSimilarity:
func = "rd_cosine"
pkl = probeFp.ToBitString()
extraFields = "%s(%s,%s) as tani" % (func, DbModule.placeHolder, details.fpColName)
cmd = _ConstructSQL(details, extraFields=extraFields)
if details.doThreshold:
# we need to do a subquery here:
cmd = "select * from (%s) tmp where tani>%f" % (cmd, details.screenThresh)
cmd += " order by tani desc"
if not details.doThreshold and details.topN > 0:
cmd += " limit %d" % details.topN
curs = conn.GetCursor()
curs.execute(cmd, (pkl,))
res = curs.fetchall()
return res
def _ConnectToDatabase(details) -> DbConnect:
if details.dbName and details.tableName:
try:
conn = DbConnect(details.dbName, details.tableName)
if hasattr(details, 'dbUser'):
conn.user = details.dbUser
if hasattr(details, 'dbPassword'):
conn.password = details.dbPassword
return conn
except Exception:
import traceback
FingerprintMols.error(f'Error: Problems establishing connection to '
f'database:{details.dbName}|{details.tableName}\n')
traceback.print_exc()
return None
def ScreenFingerprints(details, data, mol=None, probeFp=None):
""" Returns a list of results
"""
if probeFp is None:
try:
probeFp = apply(FingerprintMols.FingerprintMol, (mol, ),
details.__dict__)
except:
import traceback
FingerprintMols.error('Error: problems fingerprinting molecule.\n')
traceback.print_exc()
return []
if not probeFp:
return []
res = []
if not details.doThreshold and details.topN > 0:
topN = TopNContainer(details.topN)
else:
topN = []
res = []
count = 0
for pt in data:
fp1 = probeFp
if not details.noPickle:
if type(pt) in (types.TupleType, types.ListType):
id, fp = pt
else:
fp = pt
id = pt._fieldsFromDb[0]
score = DataStructs.FingerprintSimilarity(fp1, fp, details.metric)
else:
id, pkl = pt
score = details.metric(fp1, str(pkl))
if topN:
topN.Insert(score, id)
elif not details.doThreshold or \
(details.doThreshold and score>=details.screenThresh):
res.append((id, score))
count += 1
if hasattr(details, 'stopAfter') and count >= details.stopAfter:
break
for score, id in topN:
res.append((id, score))
return res
def ScreenFingerprints(details,data,mol=None,probeFp=None):
""" Returns a list of results
"""
if probeFp is None:
try:
probeFp = apply(FingerprintMols.FingerprintMol,(mol,),details.__dict__)
except:
import traceback
FingerprintMols.error('Error: problems fingerprinting molecule.\n')
traceback.print_exc()
return []
if not probeFp:
return []
res = []
if not details.doThreshold and details.topN>0:
topN = TopNContainer(details.topN)
else:
topN = []
res = []
count = 0
for pt in data:
fp1 = probeFp
if not details.noPickle:
if type(pt) in (types.TupleType,types.ListType):
id,fp = pt
else:
fp = pt
id = pt._fieldsFromDb[0]
score = DataStructs.FingerprintSimilarity(fp1,fp,details.metric)
else:
id,pkl = pt
score = details.metric(fp1,str(pkl))
if topN:
topN.Insert(score,id)
elif not details.doThreshold or \
(details.doThreshold and score>=details.screenThresh):
res.append((id,score))
count += 1
if hasattr(details,'stopAfter') and count >= details.stopAfter:
break
for score,id in topN:
res.append((id,score))
return res
def ScreenFingerprints(details, data, mol=None, probeFp=None):
""" Returns a list of results
"""
if probeFp is None:
try:
probeFp = FingerprintMols.FingerprintMol(mol, **details.__dict__)
except Exception:
import traceback
FingerprintMols.error('Error: problems fingerprinting molecule.\n')
traceback.print_exc()
return []
if not probeFp:
return []
if not details.doThreshold and details.topN > 0:
topN = TopNContainer(details.topN)
else:
topN = []
res = []
count = 0
for pt in data:
fp1 = probeFp
if not details.noPickle:
if isinstance(pt, (tuple, list)):
ID, fp = pt
else:
fp = pt
ID = pt._fieldsFromDb[0]
score = DataStructs.FingerprintSimilarity(fp1, fp, details.metric)
else:
ID, pkl = pt
score = details.metric(fp1, str(pkl))
if topN:
topN.Insert(score, ID)
elif not details.doThreshold or (details.doThreshold and score >= details.screenThresh):
res.append((ID, score))
count += 1
if hasattr(details, 'stopAfter') and count >= details.stopAfter:
break
for score, ID in topN:
res.append((ID, score))
return res
def ScreenFromDetails(details, mol=None):
""" Returns a list of results
"""
if not mol:
if not details.probeMol:
smi = details.probeSmiles
try:
mol = Chem.MolFromSmiles(smi)
except:
import traceback
FingerprintMols.error(
'Error: problems generating molecule for smiles: %s\n' %
(smi))
traceback.print_exc()
return
else:
mol = details.probeMol
if not mol:
return
if details.outFileName:
try:
outF = open(details.outFileName, 'w+')
except IOError:
FingerprintMols.error(
"Error: could not open output file %s for writing\n" %
(details.outFileName))
return None
else:
outF = None
if not hasattr(details, 'useDbSimilarity') or not details.useDbSimilarity:
data = GetFingerprints(details)
res = ScreenFingerprints(details, data, mol)
else:
res = ScreenInDb(details, mol)
if outF:
for pt in res:
outF.write(','.join([str(x) for x in pt]))
outF.write('\n')
return res
def GetFingerprints(details):
""" returns an iterable sequence of fingerprints
each fingerprint will have a _fieldsFromDb member whose first entry is
the id.
"""
if details.dbName and details.tableName:
conn: DbConnect = _ConnectToDatabase(details)
cmd = _ConstructSQL(details, extraFields=details.fpColName)
curs = conn.GetCursor()
# curs.execute(cmd)
# print 'CURSOR:',curs,curs.closed
if _dataSeq:
suppl = _dataSeq(curs, cmd, depickle=not details.noPickle, klass=DataStructs.ExplicitBitVect)
_dataSeq._conn = conn
return suppl
return DbFpSupplier.ForwardDbFpSupplier(data, fpColName=details.fpColName)
if details.inFileName:
try:
inF = open(details.inFileName, 'r')
except IOError:
import traceback
FingerprintMols.error(f'Error: Problems reading from file {details.inFileName}\n')
traceback.print_exc()
suppl = []
done = 0
while not done:
try:
ID, fp = pickle.load(inF)
except Exception:
done = 1
else:
fp._fieldsFromDb = [ID]
suppl.append(fp)
return suppl
return None
def ScreenFromDetails(details, mol=None):
""" Returns a list of results
"""
if not mol:
if not details.probeMol:
smi = details.probeSmiles
try:
mol = Chem.MolFromSmiles(smi)
except Exception:
import traceback
FingerprintMols.error('Error: problems generating molecule for smiles: %s\n' % (smi))
traceback.print_exc()
return
else:
mol = details.probeMol
if not mol:
return
if details.outFileName:
try:
outF = open(details.outFileName, 'w+')
except IOError:
FingerprintMols.error("Error: could not open output file %s for writing\n" %
(details.outFileName))
return None
else:
outF = None
if not hasattr(details, 'useDbSimilarity') or not details.useDbSimilarity:
data = GetFingerprints(details)
res = ScreenFingerprints(details, data, mol)
else:
res = ScreenInDb(details, mol)
if outF:
for pt in res:
outF.write(','.join([str(x) for x in pt]))
outF.write('\n')
return res