def get_descriptors(df):
PandasTools.ChangeMoleculeRendering(renderer='String')
Lmol = df['ROMol']
Ldescriptors = []
for m in Lmol:
# Calculer les propriétés chimiques
MW = round(Descriptors.ExactMolWt(m), 1)
LogP = round(Descriptors.MolLogP(m), 1)
TPSA = round(Descriptors.TPSA(m), 1)
LabuteASA = round(Descriptors.LabuteASA(m), 1)
HBA = Descriptors.NumHAcceptors(m)
HBD = Descriptors.NumHDonors(m)
FCSP3 = Lipinski.FractionCSP3(m)
MQN8 = rdMolDescriptors.MQNs_(m)[7]
MQN10 = rdMolDescriptors.MQNs_(m)[9]
NAR = Lipinski.NumAromaticRings(m)
NRB = Chem.Descriptors.NumRotatableBonds(m)
Ldescriptors.append([
MW, LogP, TPSA, LabuteASA, HBA, HBD, FCSP3, MQN8, MQN10, NAR, NRB
])
# Create pandas row for conditions results with values and information whether rule of five is violated
prop_df = pd.DataFrame(Ldescriptors)
prop_df.columns = [
'MW', 'LogP', 'TPSA', 'LabuteASA', 'HBA', 'HBD', 'FCSP3', 'MQN8',
'MQN10', 'NAR', 'NRB'
]
prop_df = prop_df.set_index(df.index)
return prop_df
def _MolQuantumNumbers(m):
'''
MQN : (molecular quantum numbers) Nguyen et al. ChemMedChem 4:1803-5 (2009)
'''
X = rdMolDescriptors.MQNs_(m)
return dict(zip(_MQN_NAMES, X))
def testMQNDetails(self):
refFile = os.path.join(RDConfig.RDCodeDir, 'Chem', 'test_data',
'MQNs_regress.pkl')
refFile2 = os.path.join(RDConfig.RDCodeDir, 'Chem', 'test_data',
'MQNs_non_strict_regress.pkl')
# figure out which definition we are currently using
m = Chem.MolFromSmiles("CC(C)(C)c1cc(O)c(cc1O)C(C)(C)C")
if Lipinski.NumRotatableBonds(m) == 2:
refFile = refFile2
with open(refFile, 'r') as intf:
buf = intf.read().replace('\r\n', '\n').encode('utf-8')
intf.close()
with io.BytesIO(buf) as inf:
pkl = inf.read()
refData = cPickle.loads(pkl, encoding='bytes')
fn = os.path.join(RDConfig.RDCodeDir, 'Chem', 'test_data',
'aromat_regress.txt')
ms = [x for x in Chem.SmilesMolSupplier(fn, delimiter='\t')]
refData2 = []
for i, m in enumerate(ms):
mqns = rdMolDescriptors.MQNs_(m)
refData2.append((m, mqns))
if mqns != refData[i][1]:
indices = [
(j, x, y)
for j, x, y in zip(range(len(mqns)), mqns, refData[i][1])
if x != y
]
print(i, Chem.MolToSmiles(m), indices)
self.assertEqual(mqns, refData[i][1])
def testMQNDetails(self):
refFile = os.path.join(os.path.dirname(__file__), 'test_data',
'MQNs_regress.pkl')
refFile2 = os.path.join(os.path.dirname(__file__), 'test_data',
'MQNs_non_strict_regress.pkl')
# figure out which definition we are currently using
m = Chem.MolFromSmiles("CC(C)(C)c1cc(O)c(cc1O)C(C)(C)C")
if Lipinski.NumRotatableBonds(m) == 2:
refFile = refFile2
with open(refFile, 'rb') as intf:
refData = pickle.load(intf)
fn = os.path.join(os.path.dirname(__file__), 'test_data',
'aromat_regress.txt')
ms = [x for x in Chem.SmilesMolSupplier(fn, delimiter='\t')]
for i, m in enumerate(ms):
mqns = rdMolDescriptors.MQNs_(m)
if mqns != refData[i][1]:
indices = [
(j, x, y)
for j, x, y in zip(range(len(mqns)), mqns, refData[i][1])
if x != y
]
print(i, Chem.MolToSmiles(m), indices)
self.assertEqual(mqns, refData[i][1])
def testMQN(self):
m = Chem.MolFromSmiles("CC(C)(C)c1cc(O)c(cc1O)C(C)(C)C")
if Lipinski.NumRotatableBonds(m) == 2:
tgt = [
42917, 274, 870, 621, 135, 1582, 29, 3147, 5463, 6999, 470,
62588, 19055, 4424, 309, 24061, 17820, 1, 9303, 24146, 16076,
5560, 4262, 646, 746, 13725, 5430, 2629, 362, 24211, 15939,
292, 41, 20, 1852, 5642, 31, 9, 1, 2, 3060, 1750
]
else:
tgt = [
42917, 274, 870, 621, 135, 1582, 29, 3147, 5463, 6999, 470,
62588, 19055, 4424, 309, 24061, 17820, 1, 8314, 24146, 16076,
5560, 4262, 646, 746, 13725, 5430, 2629, 362, 24211, 15939,
292, 41, 20, 1852, 5642, 31, 9, 1, 2, 3060, 1750
]
tgt = [
42917, 274, 870, 621, 135, 1582, 29, 3147, 5463, 6999, 470,
62588, 19055, 4424, 309, 24059, 17822, 1, 8314, 24146, 16076,
5560, 4262, 646, 746, 13725, 5430, 2629, 362, 24211, 15939,
292, 41, 20, 1852, 5642, 31, 9, 1, 2, 3060, 1750
]
fn = os.path.join(os.path.dirname(__file__), 'test_data',
'aromat_regress.txt')
ms = [x for x in Chem.SmilesMolSupplier(fn, delimiter='\t')]
vs = np.zeros((42, ), np.int32)
for m in ms:
vs += rdMolDescriptors.MQNs_(m)
self.assertEqual(list(vs), tgt)
def get_MQNs(x, from_smiles):
if from_smiles:
mol = Chem.MolFromSmiles(x)
else:
mol = x
if (mol is None) or (len(mol.GetAtoms()) == 0):
return [0] * 42
else:
return rdMolDescriptors.MQNs_(mol)
def _calculateDescriptors(mol):
df = pd.DataFrame(index=[0])
df["SlogP"] = rdMolDescriptors.CalcCrippenDescriptors(mol)[0]
df["SMR"] = rdMolDescriptors.CalcCrippenDescriptors(mol)[1]
df["LabuteASA"] = rdMolDescriptors.CalcLabuteASA(mol)
df["TPSA"] = Descriptors.TPSA(mol)
df["AMW"] = Descriptors.MolWt(mol)
df["ExactMW"] = rdMolDescriptors.CalcExactMolWt(mol)
df["NumLipinskiHBA"] = rdMolDescriptors.CalcNumLipinskiHBA(mol)
df["NumLipinskiHBD"] = rdMolDescriptors.CalcNumLipinskiHBD(mol)
df["NumRotatableBonds"] = rdMolDescriptors.CalcNumRotatableBonds(mol)
df["NumHBD"] = rdMolDescriptors.CalcNumHBD(mol)
df["NumHBA"] = rdMolDescriptors.CalcNumHBA(mol)
df["NumAmideBonds"] = rdMolDescriptors.CalcNumAmideBonds(mol)
df["NumHeteroAtoms"] = rdMolDescriptors.CalcNumHeteroatoms(mol)
df["NumHeavyAtoms"] = Chem.rdchem.Mol.GetNumHeavyAtoms(mol)
df["NumAtoms"] = Chem.rdchem.Mol.GetNumAtoms(mol)
df["NumRings"] = rdMolDescriptors.CalcNumRings(mol)
df["NumAromaticRings"] = rdMolDescriptors.CalcNumAromaticRings(mol)
df["NumSaturatedRings"] = rdMolDescriptors.CalcNumSaturatedRings(mol)
df["NumAliphaticRings"] = rdMolDescriptors.CalcNumAliphaticRings(mol)
df["NumAromaticHeterocycles"] = \
rdMolDescriptors.CalcNumAromaticHeterocycles(mol)
df["NumSaturatedHeterocycles"] = \
rdMolDescriptors.CalcNumSaturatedHeterocycles(mol)
df["NumAliphaticHeterocycles"] = \
rdMolDescriptors.CalcNumAliphaticHeterocycles(mol)
df["NumAromaticCarbocycles"] = \
rdMolDescriptors.CalcNumAromaticCarbocycles(mol)
df["NumSaturatedCarbocycles"] = \
rdMolDescriptors.CalcNumSaturatedCarbocycles(mol)
df["NumAliphaticCarbocycles"] = \
rdMolDescriptors.CalcNumAliphaticCarbocycles(mol)
df["FractionCSP3"] = rdMolDescriptors.CalcFractionCSP3(mol)
df["Chi0v"] = rdMolDescriptors.CalcChi0v(mol)
df["Chi1v"] = rdMolDescriptors.CalcChi1v(mol)
df["Chi2v"] = rdMolDescriptors.CalcChi2v(mol)
df["Chi3v"] = rdMolDescriptors.CalcChi3v(mol)
df["Chi4v"] = rdMolDescriptors.CalcChi4v(mol)
df["Chi1n"] = rdMolDescriptors.CalcChi1n(mol)
df["Chi2n"] = rdMolDescriptors.CalcChi2n(mol)
df["Chi3n"] = rdMolDescriptors.CalcChi3n(mol)
df["Chi4n"] = rdMolDescriptors.CalcChi4n(mol)
df["HallKierAlpha"] = rdMolDescriptors.CalcHallKierAlpha(mol)
df["kappa1"] = rdMolDescriptors.CalcKappa1(mol)
df["kappa2"] = rdMolDescriptors.CalcKappa2(mol)
df["kappa3"] = rdMolDescriptors.CalcKappa3(mol)
slogp_VSA = list(map(lambda i: "slogp_VSA" + str(i), list(range(1, 13))))
df = df.assign(**dict(zip(slogp_VSA, rdMolDescriptors.SlogP_VSA_(mol))))
smr_VSA = list(map(lambda i: "smr_VSA" + str(i), list(range(1, 11))))
df = df.assign(**dict(zip(smr_VSA, rdMolDescriptors.SMR_VSA_(mol))))
peoe_VSA = list(map(lambda i: "peoe_VSA" + str(i), list(range(1, 15))))
df = df.assign(**dict(zip(peoe_VSA, rdMolDescriptors.PEOE_VSA_(mol))))
MQNs = list(map(lambda i: "MQN" + str(i), list(range(1, 43))))
df = df.assign(**dict(zip(MQNs, rdMolDescriptors.MQNs_(mol))))
return df
def testMQNDetails(self):
refFile = os.path.join(RDConfig.RDCodeDir,'Chem','test_data','MQNs_regress.pkl')
refData = cPickle.load(file(refFile))
fn = os.path.join(RDConfig.RDCodeDir,'Chem','test_data','aromat_regress.txt')
ms = [x for x in Chem.SmilesMolSupplier(fn,delimiter='\t')]
for i,m in enumerate(ms):
mqns = rdMolDescriptors.MQNs_(m)
if mqns!=refData[i][1]:
indices=[(j,x,y) for j,x,y in zip(range(len(mqns)),mqns,refData[i][1]) if x!=y]
print Chem.MolToSmiles(m),indices
self.failUnlessEqual(mqns,refData[i][1])
def testMQN(self):
tgt = np.array([42917, 274, 870, 621, 135, 1582, 29, 3147, 5463,
6999, 470, 81, 19055, 4424, 309, 24061, 17820, 1,
9303, 24146, 16076, 5560, 4262, 646, 746, 13725, 5430,
2629, 362, 24211, 15939, 292, 41, 20, 1852, 5642,
31, 9, 1, 2, 3060, 1750])
fn = os.path.join(RDConfig.RDCodeDir,'Chem','test_data','aromat_regress.txt')
ms = [x for x in Chem.SmilesMolSupplier(fn,delimiter='\t')]
vs = np.zeros((42,),np.int32)
for m in ms:
vs += rdMolDescriptors.MQNs_(m)
self.failIf(False in (vs==tgt))
def get_fingerprint(in_smiles):
results = []
smiles = []
for s in in_smiles:
mol = AllChem.MolFromSmiles(s)
if mol is not None:
if mol.GetNumAtoms() > 100:
continue
mqn = Descriptors.MQNs_(mol)
results.append(np.array(mqn))
smiles.append(s)
return results, smiles
def testMQNDetails(self):
refFile = os.path.join(RDConfig.RDCodeDir,'Chem','test_data','MQNs_regress.pkl')
with open(refFile,'rb') as inf:
pkl = inf.read()
refData = cPickle.loads(pkl,encoding='bytes')
fn = os.path.join(RDConfig.RDCodeDir,'Chem','test_data','aromat_regress.txt')
ms = [x for x in Chem.SmilesMolSupplier(fn,delimiter='\t')]
for i,m in enumerate(ms):
mqns = rdMolDescriptors.MQNs_(m)
if mqns!=refData[i][1]:
indices=[(j,x,y) for j,x,y in zip(range(len(mqns)),mqns,refData[i][1]) if x!=y]
print(Chem.MolToSmiles(m),indices)
self.assertEqual(mqns,refData[i][1])
def compute_MQN_descriptors(self):
"""compute the MQN-type descriptors.
Ref: Nguyen et al. ChemMedChem 4:1803-5 (2009)
Returns:
MOE_dict: MQN dictionary, data type: int
"""
assert type(self.Molecule) == Chem.rdchem.Mol
MQN_names = []
for i in range(1, 43):
MQN_names.append('MQN' + str(i))
MQN_dict = dict(zip(MQN_names, rdDesc.MQNs_(self.Molecule)))
return MQN_dict
def feature_fp(smiles):
mol = Chem.MolFromSmiles(smiles)
fp = rdMolDescriptors.MQNs_(mol)
fp.append(rdMolDescriptors.CalcNumRotatableBonds(mol))
fp.append(rdMolDescriptors.CalcExactMolWt(mol))
fp.append(rdMolDescriptors.CalcNumRotatableBonds(mol))
fp.append(rdMolDescriptors.CalcFractionCSP3(mol))
fp.append(rdMolDescriptors.CalcNumAliphaticCarbocycles(mol))
fp.append(rdMolDescriptors.CalcNumAliphaticHeterocycles(mol))
fp.append(rdMolDescriptors.CalcNumAliphaticRings((mol)))
fp.append(rdMolDescriptors.CalcNumAromaticCarbocycles(mol))
fp.append(rdMolDescriptors.CalcNumAromaticHeterocycles(mol))
fp.append(rdMolDescriptors.CalcNumAromaticRings(mol))
fp.append(rdMolDescriptors.CalcNumBridgeheadAtoms(mol))
fp.append(rdMolDescriptors.CalcNumRings(mol))
fp.append(rdMolDescriptors.CalcNumAmideBonds(mol))
fp.append(rdMolDescriptors.CalcNumHeterocycles(mol))
fp.append(rdMolDescriptors.CalcNumSpiroAtoms(mol))
fp.append(rdMolDescriptors.CalcTPSA(mol))
return np.array(fp)
def CalculateStandAloneDescriptor(molObject):
"""
Get all standaloneDescriptor
Args:
Returns:
List
Raise:
Exceptions
"""
value_list = []
if AllChem.ComputeGasteigerCharges(molObject) == None:
value_list.append(0.0)
else:
value_list.append(1.0)
value_list.append(rdMolDescriptors.CalcNumAmideBonds(molObject))
value_list.append(rdMolDescriptors.CalcNumSpiroAtoms(molObject))
value_list.append(rdMolDescriptors.CalcNumBridgeheadAtoms(molObject))
value_list += rdMolDescriptors.MQNs_(molObject)
return value_list
#structure
embed_fn = np.nan_to_num(fngroups.values)
embed_graph = graph.values
#molecular fingerprint
#https://www.rdkit.org/UGM/2012/Landrum_RDKit_UGM.Fingerprints.Final.pptx.pdf
finger_mqn = []
finger_morgan = []
finger_maccs = []
finger_ap = []
for i in smiles:
mol = AllChem.MolFromSmiles(i)
finger_mqn.append(np.array(Descriptors.MQNs_(mol)))
finger_maccs.append(np.array(Descriptors.GetMACCSKeysFingerprint((mol))))
#finger_morgan.append(np.array(Descriptors.GetMorganFingerprint((mol))))
finger_ap.append(np.array(Descriptors.GetAtomPairFingerprint((mol))))
###
names = 'vec_spec,vec_smiles,embed_fn,finger_mqn,finger_maccs,finger_ap,embed_graph'.split(
',')
data = [
vec_spec, vec_smiles, embed_fn, finger_mqn, finger_maccs, finger_ap,
embed_graph
]
counter = 0
for i in data:
try:
res = do_pca(i)
def getMQN(mol):
l_MQNs = rdMolDescriptors.MQNs_(mol)
return l_MQNs
