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 compute_MOE_descriptors(self):
"""compute the MOE-type descriptors.
Ref:???
Returns:
MOE_dict: MOE dictionary, data type: float
"""
assert type(self.Molecule) == Chem.rdchem.Mol
MOE_dict = {}
SlogP_VSA_names = []
for i in range(1, 13):
SlogP_VSA_names.append('SlogP_VSA' + str(i))
MOE_dict.update(
dict(zip(SlogP_VSA_names, rdDesc.SlogP_VSA_(self.Molecule))))
SMR_VSA_names = []
for i in range(1, 11):
SMR_VSA_names.append('SMR_VSA' + str(i))
MOE_dict.update(
dict(zip(SMR_VSA_names, rdDesc.SMR_VSA_(self.Molecule))))
PEOE_VSA_names = []
for i in range(1, 15):
PEOE_VSA_names.append('PEOE_VSA' + str(i))
MOE_dict.update(
dict(zip(PEOE_VSA_names, rdDesc.PEOE_VSA_(self.Molecule))))
EState_VSA_names = []
for i in range(1, 12):
EState_VSA_names.append('EState_VSA' + str(i))
MOE_dict.update(
dict(zip(EState_VSA_names, EState.EState_VSA_(self.Molecule))))
VSA_EState_names = []
for i in range(1, 12):
VSA_EState_names.append('VSA_EState' + str(i))
MOE_dict.update(
dict(zip(VSA_EState_names, EState.VSA_EState_(self.Molecule))))
return MOE_dict
def extractFeatureData(mol):
smr_vsa = rdMolDescriptors.SMR_VSA_(mol)
slogp_vsa = rdMolDescriptors.SlogP_VSA_(mol)
peoe_vsa = rdMolDescriptors.PEOE_VSA_(mol)
hbd = rdMolDescriptors.CalcNumHBD(mol)
hba = rdMolDescriptors.CalcNumHBA(mol)
feats = [smr_vsa, slogp_vsa, peoe_vsa, hbd, hba]
feature_data = []
for f in feats:
if (isinstance(f, int)):
feature_data.append(f)
else:
for data in f:
feature_data.append(data)
#feature_data = np.asarray(feature_data) # convert to numpy array
return feature_data
def extractFeatureData(mol):
global index_of_1d_feature
smr_vsa = rdMolDescriptors.SMR_VSA_(mol)
slogp_vsa = rdMolDescriptors.SlogP_VSA_(mol)
peoe_vsa = rdMolDescriptors.PEOE_VSA_(mol)
hbd = rdMolDescriptors.CalcNumHBD(mol)
hba = rdMolDescriptors.CalcNumHBA(mol)
index_of_1d_feature = -1 # Need to make sure this references the index of a 1D feature
# (a negative index refers to counting backwards from the end of a list)
feats = [smr_vsa, slogp_vsa, peoe_vsa, hbd, hba]
feature_data = []
for f in feats:
if (isinstance(f, int)):
feature_data.append(f)
else:
for data in f:
feature_data.append(data)
#feature_data = np.asarray(feature_data) # convert to numpy array
return feature_data