def test3Reactions(self): txt = """BoronicAcid\t[$(B-!@[#6])](O)(O)\tBoronic Acid\t[#6:1][B:2]([O:3])[O:4]>>[#6:1].[B:2]([O:3])[O:4] BoronicAcid.Aromatic\t[$(B-!@c)](O)(O)\tAromatic\t[c:1][B:2]([O:3])[O:4]>>[c:1].[B:2]([O:3])[O:4] BoronicAcid.Aliphatic\t[$(B-!@C)](O)(O)\tAliphatic\t[C:1][B:2]([O:3])[O:4]>>[C:1].[B:2]([O:3])[O:4] """ hierarchy = FunctionalGroups.BuildFuncGroupHierarchy(data=txt) self.assertTrue(hierarchy) self.assertEqual(len(hierarchy), 1) self.assertEqual(len(hierarchy[0].children), 2) self.assertNotEqual(hierarchy[0].rxnSmarts, '') self.assertNotEqual(hierarchy[0].children[0].rxnSmarts, '')
def test3Reactions(self):
txt = """BoronicAcid\t[$(B-!@[#6])](O)(O)\tBoronic Acid\t[#6:1][B:2]([O:3])[O:4]>>[#6:1].[B:2]([O:3])[O:4]
BoronicAcid.Aromatic\t[$(B-!@c)](O)(O)\tAromatic\t[c:1][B:2]([O:3])[O:4]>>[c:1].[B:2]([O:3])[O:4]
BoronicAcid.Aliphatic\t[$(B-!@C)](O)(O)\tAliphatic\t[C:1][B:2]([O:3])[O:4]>>[C:1].[B:2]([O:3])[O:4]
"""
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy(data=txt)
self.failUnless(hierarchy)
self.failUnless(len(hierarchy) == 1)
self.failUnless(len(hierarchy[0].children) == 2)
self.failUnless(hierarchy[0].rxnSmarts != '')
self.failUnless(hierarchy[0].children[0].rxnSmarts != '')
def test2Comments(self):
txt = """
AcidChloride\tC(=O)Cl\tAcid Chloride
AcidChloride.Benzoyl\tC(=O)(Cl)c1ccccc1\tBenzoyl
Amine\tN\tAmine
Amine.Primary\t[N;H2]\tPrimary
//Amine.Primary.Aromatic\t[N;H2][a]\tPrimary Aromatic
Amine.Aromatic\tN[a]\tAromatic
"""
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy(data=txt)
self.assertTrue(hierarchy)
self.assertEqual(len(hierarchy), 2)
self.assertEqual(len(hierarchy[0]), 2)
self.assertEqual(len(hierarchy[1]), 3)
def test4Hs(self):
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy()
inName = os.path.join(RDConfig.RDCodeDir, 'Chem', 'test_data', 'NCI_5K_TPSA.csv')
with open(inName, 'r') as inF:
ms = [Chem.MolFromSmiles(x.split(',')[0]) for x in inF if x[0] != '#']
for m in ms:
mh = Chem.AddHs(m)
fp = FunctionalGroups.CreateMolFingerprint(m, hierarchy)
fph = FunctionalGroups.CreateMolFingerprint(mh, hierarchy)
if fp != fph:
print(Chem.MolToSmiles(m))
print(fp.ToBitString())
print(fph.ToBitString())
self.assertEqual(fp, fph)
def test4Hs(self):
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy()
inName = os.path.join(RDConfig.RDCodeDir, 'Chem', 'test_data',
'NCI_5K_TPSA.csv')
inF = open(inName, 'r')
lines = inF.readlines()
ms = [
Chem.MolFromSmiles(x.split(',')[0]) for x in lines if x[0] != '#'
]
for m in ms:
mh = Chem.AddHs(m)
fp = FunctionalGroups.CreateMolFingerprint(m, hierarchy)
fph = FunctionalGroups.CreateMolFingerprint(mh, hierarchy)
if fp != fph:
print fp.ToBitString()
print fph.ToBitString()
self.failUnlessEqual(fp, fph)
def RetrieveFunctionalGroupsInfo():
"""Retrieve functional groups information"""
MiscUtil.PrintInfo(
"\nRetrieving data from default RDKit functional groups hierarchy file Functional_Group_Hierarchy.txt..."
)
FunctionalGroupNamesFile = OptionsInfo["GroupNamesFile"]
FunctionalGroupsNodes = FunctionalGroups.BuildFuncGroupHierarchy(
FunctionalGroupNamesFile)
FunctionalGroupsMap['Names'] = []
FunctionalGroupsMap['SMARTSPattern'] = {}
RetrieveDataFromFunctionalGroupsHierarchy(FunctionalGroupsNodes)
if not len(FunctionalGroupsMap['Names']):
MiscUtil.PrintError(
"Failed to retrieve any functional group names and SMARTS patterns..."
)
MiscUtil.PrintInfo(
"Total number of functional groups present functional group hierarchy: %d"
% (len(FunctionalGroupsMap['Names'])))
from rdkit.Chem import Draw
from rdkit.Chem import FunctionalGroups
from time import time
import csv
start = time()
# open target csv file
csvfile = open("fg_smils.csv", 'a')
header = ["pubchemId","functionalGroup", "SMILES"]
writer = csv.writer(csvfile)
writer.writerow(header)
smilesfile = "CID-SMILES"
fgs = FunctionalGroups.BuildFuncGroupHierarchy()
#for filename in filenames:
suppl = Chem.SmilesMolSupplier(smilesfile, delimiter=",",
smilesColumn=1, nameColumn=0, titleLine=False )
mols = [x for x in suppl if x is not None]
del suppl
print("processing %s with %d valid compounds" % (filename, (len(mols))))
#Draw.MolsToGridImage(mols[:20], molsPerRow=4, legends=[x.GetProp('_Name') for x in mols])
'''
def test1Basics(self):
txt = """
AcidChloride\tC(=O)Cl\tAcid Chloride
AcidChloride.Benzoyl\tC(=O)(Cl)c1ccccc1\tBenzoyl
Amine\tN\tAmine
Amine.Primary\t[N;H2]\tPrimary
Amine.Primary.Aromatic\t[N;H2][a]\tPrimary Aromatic
Amine.Aromatic\tN[a]\tAromatic
"""
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy(data=txt)
self.assertTrue(hierarchy)
self.assertEqual(len(hierarchy), 2)
self.assertEqual(len(hierarchy[0]), 2)
self.assertEqual(len(hierarchy[1]), 4)
self.assertEqual(hierarchy[0].name, 'Acid Chloride')
self.assertEqual(hierarchy[0].children[0].name, 'Benzoyl')
self.assertEqual(hierarchy[0].label, 'AcidChloride')
self.assertEqual(hierarchy[0].rxnSmarts, '')
m = Chem.MolFromSmiles('ClC(=O)CCCNc1ccccc1')
fp = FunctionalGroups.CreateMolFingerprint(m, hierarchy)
self.assertEqual(fp, [1, 0, 1, 0, 0, 1])
m = Chem.MolFromSmiles('OC(=O)CCC')
fp = FunctionalGroups.CreateMolFingerprint(m, hierarchy)
self.assertEqual(fp, [0, 0, 0, 0, 0, 0])
# make sure we get the same hierarchy on the second call:
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy(data=txt)
self.assertTrue(hierarchy)
self.assertEqual(len(hierarchy), 2)
self.assertEqual(len(hierarchy[0]), 2)
self.assertEqual(len(hierarchy[1]), 4)
self.assertEqual(hierarchy[0].name, 'Acid Chloride')
self.assertEqual(hierarchy[0].children[0].name, 'Benzoyl')
self.assertEqual(hierarchy[0].label, 'AcidChloride')
self.assertEqual(hierarchy[0].rxnSmarts, '')
# if we edit this hierarchy it doesn't affect the global one:
hierarchy.pop(0)
self.assertEqual(len(hierarchy[0]), 4)
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy(data=txt)
self.assertTrue(hierarchy)
self.assertEqual(len(hierarchy), 2)
self.assertEqual(len(hierarchy[0]), 2)
self.assertEqual(len(hierarchy[1]), 4)
self.assertEqual(hierarchy[0].name, 'Acid Chloride')
self.assertEqual(hierarchy[0].children[0].name, 'Benzoyl')
self.assertEqual(hierarchy[0].label, 'AcidChloride')
self.assertEqual(hierarchy[0].rxnSmarts, '')
# and if we edit the global one and don't force, we get the edited one:
FunctionalGroups.hierarchy.pop(0)
self.assertEqual(len(FunctionalGroups.hierarchy[0]), 4)
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy(data=txt)
self.assertTrue(hierarchy)
self.assertEqual(len(hierarchy), 1)
self.assertEqual(len(hierarchy[0]), 4)
# but a force gets us back:
hierarchy = FunctionalGroups.BuildFuncGroupHierarchy(data=txt, force=True)
self.assertEqual(len(hierarchy), 2)
self.assertEqual(len(hierarchy[0]), 2)
self.assertEqual(len(hierarchy[1]), 4)