def atom_level_descriptors(mol,
include=['functional'],
asOneHot=False,
ORIGINAL_VERSION=False):
'''
Given an RDKit mol, returns an N_atom-long list of lists,
each of which contains atom-level descriptors and their names
returns: (label, attributes)
'''
attributes = [[] for i in mol.GetAtoms()]
labels = []
if 'functional' in include:
[attributes[i].append(x[0]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(mol))]
labels.append('Crippen contribution to logp')
[attributes[i].append(x[1]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(mol))]
labels.append('Crippen contribution to mr')
[attributes[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcTPSAContribs(mol))]
labels.append('TPSA contribution')
[attributes[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcLabuteASAContribs(mol)[0])]
labels.append('Labute ASA contribution')
[attributes[i].append(x) \
for (i, x) in enumerate(EState.EStateIndices(mol))]
labels.append('EState Index')
rdPartialCharges.ComputeGasteigerCharges(mol)
[attributes[i].append(float(a.GetProp('_GasteigerCharge'))) \
for (i, a) in enumerate(mol.GetAtoms())]
labels.append('Gasteiger partial charge')
# Gasteiger partial charges sometimes gives NaN
for i in range(len(attributes)):
if np.isnan(attributes[i][-1]):
attributes[i][-1] = 0.0
[attributes[i].append(float(a.GetProp('_GasteigerHCharge'))) \
for (i, a) in enumerate(mol.GetAtoms())]
labels.append('Gasteiger hydrogen partial charge')
# Gasteiger partial charges sometimes gives NaN
for i in range(len(attributes)):
if np.isnan(attributes[i][-1]):
attributes[i][-1] = 0.0
if 'structural' in include:
[attributes[i].extend(atom_structural(mol.GetAtomWithIdx(i), asOneHot = asOneHot, ORIGINAL_VERSION = ORIGINAL_VERSION)) \
for i in range(len(attributes))]
labels.append('--many structural--')
return (labels, attributes)
def testCrippen(self):
mol = Chem.MolFromSmiles("n1ccccc1CO");
contribs = rdMD._CalcCrippenContribs(mol)
self.failUnlessEqual(len(contribs),mol.GetNumAtoms());
ts = [0]*mol.GetNumAtoms()
contribs = rdMD._CalcCrippenContribs(mol,force=True,atomTypes=ts)
self.failUnlessEqual(ts,[59, 25, 25, 25, 25, 28, 17, 69])
ls = ['']*mol.GetNumAtoms()
contribs = rdMD._CalcCrippenContribs(mol,force=True,atomTypeLabels=ls)
self.failUnlessEqual(ls,['N11', 'C18', 'C18', 'C18', 'C18', 'C21', 'C10', 'O2'])
def testCrippen(self):
mol = Chem.MolFromSmiles("n1ccccc1CO");
contribs = rdMD._CalcCrippenContribs(mol)
self.assertEqual(len(contribs),mol.GetNumAtoms());
ts = [0]*mol.GetNumAtoms()
contribs = rdMD._CalcCrippenContribs(mol,force=True,atomTypes=ts)
self.assertEqual(ts,[59, 25, 25, 25, 25, 28, 17, 69])
ls = ['']*mol.GetNumAtoms()
contribs = rdMD._CalcCrippenContribs(mol,force=True,atomTypeLabels=ls)
self.assertEqual(ls,['N11', 'C18', 'C18', 'C18', 'C18', 'C21', 'C10', 'O2'])
def mol_to_nx(mol) -> nx.Graph:
G = nx.Graph()
conf = mol.GetConformer()
SanitizeMol(mol,
SanitizeFlags.SANITIZE_ALL ^ SanitizeFlags.SANITIZE_PROPERTIES)
ComputeGasteigerCharges(mol)
ring_info = mol.GetRingInfo()
crippen_contribs = rdMolDescriptors._CalcCrippenContribs(mol)
tpsa_contribs = rdMolDescriptors._CalcTPSAContribs(mol)
for atom in mol.GetAtoms():
idx = atom.GetIdx()
# if atom.GetSymbol() == 'N' and atom.GetTotalValence() == 2:
# formal_charge = -1
# elif atom.GetSymbol() == 'N' and atom.GetTotalValence() == 4:
# formal_charge = 1
# elif atom.GetSymbol() == 'O' and atom.GetTotalValence() == 1:
# formal_charge = -1
# else:
# formal_charge = atom.GetFormalCharge()
formal_charge = atom.GetFormalCharge()
G.add_node(
idx,
pos=conf.GetAtomPosition(idx),
formal_charge=formal_charge,
chiral_tag=atom.GetChiralTag(),
hybridization=atom.GetHybridization(),
# num_explicit_hs=atom.GetNumExplicitHs(), # All same
is_aromatic=atom.GetIsAromatic(),
num_atom_rings=ring_info.NumAtomRings(idx),
is_in_ring_size3=atom.IsInRingSize(3),
is_in_ring_size4=atom.IsInRingSize(4),
is_in_ring_size5=atom.IsInRingSize(5),
is_in_ring_size6=atom.IsInRingSize(6),
symbol=atom.GetSymbol(),
total_valence=atom.GetTotalValence(),
gasteiger_charge=atom.GetProp('_GasteigerCharge'),
num_implicit_hs=atom.GetNumImplicitHs(),
total_degree=atom.GetTotalDegree(),
crippen_logp=crippen_contribs[idx][0],
crippen_mr=crippen_contribs[idx][1],
tpsa=tpsa_contribs[idx],
)
for bond in mol.GetBonds():
G.add_edge(
bond.GetBeginAtomIdx(),
bond.GetEndAtomIdx(),
bond_type=bond.GetBondType(),
is_conjugated=bond.GetIsConjugated(),
)
return G
def calculate_logP(self, mol):
"""Calculates Crippen contributions, i.e. logP of ligand molecule.
Takes:
* mol * - mol file in rdkit environment
Returns:
* contribs * - tuple of Wildman-Crippen logP, MR (molar refractivity -
measure of the volume occupied by a molecule of the substance) values
"""
contribs = rdMolDescriptors._CalcCrippenContribs(mol)
return contribs
def calculate_logP(self,mol): """Calculates Crippen contributions, i.e. logP of ligand molecule. Takes: * mol * - mol file in rdkit environment Returns: * contribs * - tuple of Wildman-Crippen logP, MR (molar refractivity - measure of the volume occupied by a molecule of the substance) values """ contribs = rdMolDescriptors._CalcCrippenContribs(mol) return contribs
def test8CrippenO3A(self):
sdf = os.path.join(RDConfig.RDBaseDir,'Code','GraphMol',
'MolAlign', 'test_data', 'ref_e2.sdf')
# alignedSdf = os.path.join(RDConfig.RDBaseDir,'Code','GraphMol',
# 'MolAlign', 'test_data', 'ref_e2_pyCrippenO3A.sdf')
molS = Chem.SDMolSupplier(sdf, True, False)
# molW = Chem.SDWriter(alignedSdf)
refNum = 48
refMol = molS[refNum]
cumScore = 0.0
cumMsd = 0.0
refList = rdMolDescriptors._CalcCrippenContribs(refMol, True)
for prbMol in molS:
prbList = rdMolDescriptors._CalcCrippenContribs(prbMol, True)
pyO3A = rdMolAlign.GetCrippenO3A(prbMol, refMol, prbList, refList)
cumScore += pyO3A.Score()
rmsd = pyO3A.Align()
cumMsd += rmsd * rmsd
# molW.write(prbMol)
cumMsd /= len(molS)
self.failUnlessAlmostEqual(cumScore,4918,0)
self.failUnlessAlmostEqual(math.sqrt(cumMsd),.304,3)
def test10CrippenO3A(self):
sdf = os.path.join(RDConfig.RDBaseDir,'Code','GraphMol',
'MolAlign', 'test_data', 'ref_e2.sdf')
alignedSdf = os.path.join(RDConfig.RDBaseDir,'Code','GraphMol',
'MolAlign', 'test_data', 'ref_e2_pyCrippenO3A.sdf')
molS = Chem.SDMolSupplier(sdf, True, False)
molW = Chem.SDWriter(alignedSdf)
refNum = 48
refMol = molS[refNum]
cumScore = 0.0
cumMsd = 0.0
refList = rdMolDescriptors._CalcCrippenContribs(refMol, True)
for prbMol in molS:
prbList = rdMolDescriptors._CalcCrippenContribs(prbMol, True)
pyO3A = rdMolAlign.GetCrippenO3A(prbMol, refMol, prbList, refList)
cumScore += pyO3A.Score()
rmsd = pyO3A.Align()
cumMsd += rmsd * rmsd
molW.write(prbMol)
cumMsd /= len(molS)
self.assertAlmostEqual(cumScore,4918,0)
self.assertAlmostEqual(math.sqrt(cumMsd),.304,3)
def align_set_of_ligands(ligands: Sequence) -> Tuple[List[Chem.Mol], List[float]]:
""" Align a set of ligands to each other
Parameters
----------
ligands : list of rdkit.Chem.rdchem.Mol or rdkit.Chem.SmilesMolSupplier or rdkit.Chem.SDMolSupplier
List of ligands.
Returns
----------
aligned_molecules : list of rdkit.Chem.rdchem.Mol
List of aligned ligands.
crippen_score : list of float
List with crippen scores calculated during the alignment.
"""
if not isinstance(ligands, list):
ligands = list(ligands)
molecules = copy.deepcopy(ligands)
molecules = [generate_conformers(mol, 100) for mol in molecules]
crippen_contribs = [rdMolDescriptors._CalcCrippenContribs(mol) for mol in molecules]
crippen_ref_contrib = crippen_contribs[0]
crippen_prob_contribs = crippen_contribs
ref_mol = molecules[0]
probe_mols = molecules
crippen_score = []
aligned_molecules = []
for idx, mol in enumerate(probe_mols):
tempscore = []
for cid in range(100):
crippenO3A = rdMolAlign.GetCrippenO3A(mol, ref_mol, crippen_prob_contribs[idx], crippen_ref_contrib, cid, 0)
crippenO3A.Align()
tempscore.append(crippenO3A.Score())
best = np.argmax(tempscore)
mol_string = Chem.MolToMolBlock(mol, confId=int(best))
temp_mol = Chem.MolFromMolBlock(mol_string, removeHs=False)
crippen_score.append(tempscore[best])
aligned_molecules.append(temp_mol)
return aligned_molecules, crippen_score
def assignProperties(mol):
'''
Calculate atom-level descriptors that can be used in featurization
'''
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(mol)):
mol.GetAtomWithIdx(i).SetDoubleProp('crippen_logp',x[0])
mol.GetAtomWithIdx(i).SetDoubleProp('crippen_mr', x[1])
for (i, x) in enumerate(rdMolDescriptors._CalcTPSAContribs(mol)):
mol.GetAtomWithIdx(i).SetDoubleProp('tpsa', x)
for (i, x) in enumerate(rdMolDescriptors._CalcLabuteASAContribs(mol)[0]):
mol.GetAtomWithIdx(i).SetDoubleProp('asa', x)
for (i, x) in enumerate(EState.EStateIndices(mol)):
mol.GetAtomWithIdx(i).SetDoubleProp('estate', x)
rdPartialCharges.ComputeGasteigerCharges(mol) # '_GasteigerCharge', '_GasteigerHCharge'
def get_molecular_attributes(rdmol):
"""
Molecular attributes calculated as:
[Crippen contribution to logp,
Crippen contribution to mr,
TPSA contribution,
Labute ASA contribution,
EState Index,
Gasteiger partial charge,
Gasteiger hydrogen partial charge]
Parameters
----------
rdmol : rdkit.Chem.rdchem.Mol
rdkit molecule class
Returns
-------
attributes : list
feature vector
"""
attributes = [[] for _ in rdmol.GetAtoms()]
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(rdmol)):
attributes[i].append(x[0])
attributes[i].append(x[1])
for (i, x) in enumerate(rdMolDescriptors._CalcTPSAContribs(rdmol)):
attributes[i].append(x)
for (i, x) in enumerate(rdMolDescriptors._CalcLabuteASAContribs(rdmol)[0]):
attributes[i].append(x)
for (i, x) in enumerate(EState.EStateIndices(rdmol)):
attributes[i].append(x)
rdPartialCharges.ComputeGasteigerCharges(rdmol)
for (i, a) in enumerate(rdmol.GetAtoms()):
val = float(a.GetProp('_GasteigerCharge'))
if val == val and val < np.inf:
attributes[i].append(val)
else:
attributes[i].append(0.0)
for (i, a) in enumerate(rdmol.GetAtoms()):
val = float(a.GetProp('_GasteigerHCharge'))
if val == val and val < np.inf:
attributes[i].append(val)
else:
attributes[i].append(0.0)
return attributes
def molToGraph(rdmol):
'''
Converts an RDKit molecule to an attributed undirected graph
@param rdmol: RDKit molecule
@return: Graph
'''
graph = Graph()
# Calculate atom-level molecule descriptors
nodesFeatures = [[] for i in rdmol.GetAtoms()]
#6 (25) Crippen contribution to logp
[nodesFeatures[i].append(x[0]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(rdmol))]
#7 (26) Crippen contribution to mr
[nodesFeatures[i].append(x[1]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(rdmol))]
#8 (27) TPSA contribution
[nodesFeatures[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcTPSAContribs(rdmol))]
#9 (28) Labute ASA contribution
[nodesFeatures[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcLabuteASAContribs(rdmol)[0])]
#10 (29) EState Index
[nodesFeatures[i].append(x) \
for (i, x) in enumerate(EState.EStateIndices(rdmol))]
# Calculate Gasteiger charges for features 30 and 31
rdPartialCharges.ComputeGasteigerCharges(rdmol)
# The computed charges are stored on each atom with computed property
# under the name _GasteigerCharge and _GasteigerHCharge.
# Values could be NaN.
#11 (30)
for (i, a) in enumerate(rdmol.GetAtoms()):
if np.isnan(float(a.GetProp('_GasteigerCharge'))) or np.isinf(
float(a.GetProp('_GasteigerCharge'))):
nodesFeatures[i].append(0.0)
else:
nodesFeatures[i].append(float(a.GetProp('_GasteigerCharge')))
#12 (31)
for (i, a) in enumerate(rdmol.GetAtoms()):
if np.isnan(float(a.GetProp('_GasteigerHCharge'))) or np.isinf(
float(a.GetProp('_GasteigerHCharge'))):
nodesFeatures[i].append(0.0)
else:
nodesFeatures[i].append(float(a.GetProp('_GasteigerHCharge')))
# Add edges to graph
for bond in rdmol.GetBonds():
edge = Edge()
edge.id = bond.GetIdx()
edge.features = getBondFeatures(bond).astype('float32')
edge.ends = (bond.GetBeginAtomIdx(), bond.GetEndAtomIdx())
graph.edges.append(edge)
# Add nodes to graph
for i, atom in enumerate(rdmol.GetAtoms()):
node = Node()
node.id = atom.GetIdx()
node.features = getAtomFeatures(atom, nodesFeatures[i])
for neighbor in atom.GetNeighbors():
node.neighbors.append(
(neighbor.GetIdx(),
rdmol.GetBondBetweenAtoms(atom.GetIdx(),
neighbor.GetIdx()).GetIdx()))
graph.nodes.append(node)
graph.nodeNum = len(graph.nodes)
graph.nodeFeatureDim = len(graph.nodes[0].features)
if (len(graph.edges) > 0):
graph.edgeFeatureDim = len(graph.edges[0].features)
return graph
def molToGraph(rdmol, molecular_attributes=False):
'''Converts an RDKit molecule to an attributed undirected graph'''
# Initialize
graph = Graph()
graph.molecular_attributes = molecular_attributes
# Calculate atom-level molecule descriptors
attributes = [[] for i in rdmol.GetAtoms()]
if molecular_attributes:
labels = []
[attributes[i].append(x[0]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(rdmol))]
labels.append('Crippen contribution to logp')
[attributes[i].append(x[1]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(rdmol))]
labels.append('Crippen contribution to mr')
[attributes[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcTPSAContribs(rdmol))]
labels.append('TPSA contribution')
[attributes[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcLabuteASAContribs(rdmol)[0])]
labels.append('Labute ASA contribution')
[attributes[i].append(x) \
for (i, x) in enumerate(EState.EStateIndices(rdmol))]
labels.append('EState Index')
rdPartialCharges.ComputeGasteigerCharges(rdmol)
[attributes[i].append(float(a.GetProp('_GasteigerCharge'))) \
for (i, a) in enumerate(rdmol.GetAtoms())]
labels.append('Gasteiger partial charge')
# Gasteiger partial charges sometimes gives NaN
for i in range(len(attributes)):
if np.isnan(attributes[i][-1]) or np.isinf(attributes[i][-1]):
attributes[i][-1] = 0.0
[attributes[i].append(float(a.GetProp('_GasteigerHCharge'))) \
for (i, a) in enumerate(rdmol.GetAtoms())]
labels.append('Gasteiger hydrogen partial charge')
# Gasteiger partial charges sometimes gives NaN
for i in range(len(attributes)):
if np.isnan(attributes[i][-1]) or np.isinf(attributes[i][-1]):
attributes[i][-1] = 0.0
# Add bonds
for bond in rdmol.GetBonds():
edge = Edge()
edge.i = bond.GetIdx()
edge.attributes = bondAttributes(bond)
edge.connects = (bond.GetBeginAtomIdx(), bond.GetEndAtomIdx())
graph.edges.append(edge)
# Add atoms
for k, atom in enumerate(rdmol.GetAtoms()):
node = Node()
node.i = atom.GetIdx()
node.attributes = atomAttributes(atom, extra_attributes=attributes[k])
for neighbor in atom.GetNeighbors():
node.neighbors.append(
(neighbor.GetIdx(),
rdmol.GetBondBetweenAtoms(atom.GetIdx(),
neighbor.GetIdx()).GetIdx()))
graph.nodes.append(node)
# Add counts, for convenience
graph.num_edges = len(graph.edges)
graph.num_nodes = len(graph.nodes)
return graph
def make_graph(name, gb_structure, gb_scalar_coupling):
# ['id', 'molecule_name', 'atom_index_0', 'atom_index_1', 'type','scalar_coupling_constant']
coupling_df = gb_scalar_coupling.get_group(name)
# [molecule_name,atom_index,atom,x,y,z]
df = gb_structure.get_group(name)
df = df.sort_values(['atom_index'], ascending=True)
a = df.atom.values.tolist()
xyz = df[['x', 'y', 'z']].values
mol = mol_from_axyz(a, xyz)
mol_op = openbabel.OBMol()
obConversion.ReadFile(mol_op, f'../input/champs-scalar-coupling/structures/{name}.xyz')
factory = ChemicalFeatures.BuildFeatureFactory(os.path.join(RDConfig.RDDataDir, 'BaseFeatures.fdef'))
feature = factory.GetFeaturesForMol(mol)
num_atom = mol.GetNumAtoms()
symbol = np.zeros((num_atom, len(SYMBOL)), np.uint8) # category
acceptor = np.zeros((num_atom, 1), np.uint8)
donor = np.zeros((num_atom, 1), np.uint8)
aromatic = np.zeros((num_atom, 1), np.uint8)
hybridization = np.zeros((num_atom, len(HYBRIDIZATION)), np.uint8)
num_h = np.zeros((num_atom, 1), np.float32) # real
atomic = np.zeros((num_atom, 1), np.float32)
# new features
degree = np.zeros((num_atom, 1), np.uint8)
formalCharge = np.zeros((num_atom, 1), np.float32)
chiral_tag = np.zeros((num_atom, 1), np.uint8)
crippen_contribs = np.zeros((num_atom, 2), np.float32)
tpsa = np.zeros((num_atom, 1), np.float32)
labute_asac = np.zeros((num_atom, 1), np.float32)
gasteiger_charges = np.zeros((num_atom, 1), np.float32)
esataindices = np.zeros((num_atom, 1), np.float32)
atomic_radiuss = np.zeros((num_atom, 1), np.float32)
electronegate = np.zeros((num_atom, 1), np.float32)
electronegate_sqre = np.zeros((num_atom, 1), np.float32)
mass = np.zeros((num_atom, 1), np.float32)
van = np.zeros((num_atom, 1), np.float32)
cov = np.zeros((num_atom, 1), np.float32)
ion = np.zeros((num_atom, 1), np.float32)
for i in range(num_atom):
atom = mol.GetAtomWithIdx(i)
atom_op = mol_op.GetAtomById(i)
symbol[i] = one_hot_encoding(atom.GetSymbol(), SYMBOL)
aromatic[i] = atom.GetIsAromatic()
hybridization[i] = one_hot_encoding(atom.GetHybridization(), HYBRIDIZATION)
num_h[i] = atom.GetTotalNumHs(includeNeighbors=True)
atomic[i] = atom.GetAtomicNum()
degree[i] = atom.GetTotalDegree()
formalCharge[i] = atom.GetFormalCharge()
chiral_tag[i] = int(atom.GetChiralTag())
crippen_contribs[i] = rdMolDescriptors._CalcCrippenContribs(mol)[i]
tpsa[i] = rdMolDescriptors._CalcTPSAContribs(mol)[i]
labute_asac[i] = rdMolDescriptors._CalcLabuteASAContribs(mol)[0][i]
gasteiger_charges[i] = atom_op.GetPartialCharge()
esataindices[i] = EState.EStateIndices(mol)[i]
atomic_radiuss[i] = atomic_radius[atom.GetSymbol()]
electronegate[i] = electronegativity[atom.GetSymbol()]
electronegate_sqre[i] = electronegativity_square[atom.GetSymbol()]
mass[i] = atomic_mass[atom.GetSymbol()]
van[i] = vanderwaalsradius[atom.GetSymbol()]
cov[i] = covalenzradius[atom.GetSymbol()]
ion[i] = ionization_energy[atom.GetSymbol()]
for t in range(0, len(feature)):
if feature[t].GetFamily() == 'Donor':
for i in feature[t].GetAtomIds():
donor[i] = 1
elif feature[t].GetFamily() == 'Acceptor':
for i in feature[t].GetAtomIds():
acceptor[i] = 1
num_edge = num_atom * num_atom - num_atom
edge_index = np.zeros((num_edge, 2), np.uint32)
bond_type = np.zeros((num_edge, len(BOND_TYPE)), np.uint32)
distance = np.zeros((num_edge, 1), np.float32)
angle = np.zeros((num_edge, 1), np.float32)
norm_xyz = preprocessing.normalize(xyz, norm='l2')
ij = 0
for i in range(num_atom):
for j in range(num_atom):
if i == j: continue
edge_index[ij] = [i, j]
bond = mol.GetBondBetweenAtoms(i, j)
if bond is not None:
bond_type[ij] = one_hot_encoding(bond.GetBondType(), BOND_TYPE)
distance[ij] = np.linalg.norm(xyz[i] - xyz[j])
angle[ij] = (norm_xyz[i] * norm_xyz[j]).sum()
ij += 1
xyz = xyz * 1.889726133921252
atom = System(symbols=a, positions=xyz)
acsf = ACSF_GENERATOR.create(atom)
l = []
for item in coupling_df[['atom_index_0', 'atom_index_1']].values.tolist():
i = edge_index.tolist().index(item)
l.append(i)
l = np.array(l)
coupling_edge_index = np.concatenate([coupling_df[['atom_index_0', 'atom_index_1']].values, l.reshape(len(l), 1)],
axis=1)
coupling = Coupling(coupling_df['id'].values,
coupling_df[['fc', 'sd', 'pso', 'dso']].values,
coupling_edge_index,
np.array([COUPLING_TYPE.index(t) for t in coupling_df.type.values], np.int32),
coupling_df['scalar_coupling_constant'].values,
)
graph = Graph(
name,
Chem.MolToSmiles(mol),
[a, xyz],
[acsf, symbol, acceptor, donor, aromatic, hybridization, num_h, atomic, degree, formalCharge, chiral_tag,
crippen_contribs, tpsa, labute_asac, gasteiger_charges, esataindices, atomic_radiuss, electronegate,
electronegate_sqre, mass, van, cov, ion],
[bond_type, distance, angle, ],
edge_index,
coupling,
)
return graph
def molToGraph(rdmol, bondtype_list_order, atomtype_list_order, molecular_attributes = False):
'''Converts an RDKit molecule to an attributed undirected graph'''
# Initialize
graph = Graph()
graph.molecular_attributes = molecular_attributes
graph.bondtype_list_order = bondtype_list_order
bond_list = bondtype_list_order
graph.atomtype_list_order = atomtype_list_order
# Calculate atom-level molecule descriptors
attributes = [[] for i in rdmol.GetAtoms()]
if molecular_attributes:
labels = []
[attributes[i].append(x[0]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(rdmol))]
labels.append('Crippen contribution to logp')
[attributes[i].append(x[1]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(rdmol))]
labels.append('Crippen contribution to mr')
[attributes[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcTPSAContribs(rdmol))]
labels.append('TPSA contribution')
[attributes[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcLabuteASAContribs(rdmol)[0])]
labels.append('Labute ASA contribution')
[attributes[i].append(x) \
for (i, x) in enumerate(EState.EStateIndices(rdmol))]
labels.append('EState Index')
rdPartialCharges.ComputeGasteigerCharges(rdmol)
[attributes[i].append(float(a.GetProp('_GasteigerCharge'))) \
for (i, a) in enumerate(rdmol.GetAtoms())]
labels.append('Gasteiger partial charge')
# Gasteiger partial charges sometimes gives NaN
for i in range(len(attributes)):
if np.isnan(attributes[i][-1]) or np.isinf(attributes[i][-1]):
attributes[i][-1] = 0.0
[attributes[i].append(float(a.GetProp('_GasteigerHCharge'))) \
for (i, a) in enumerate(rdmol.GetAtoms())]
labels.append('Gasteiger hydrogen partial charge')
# Gasteiger partial charges sometimes gives NaN
for i in range(len(attributes)):
if np.isnan(attributes[i][-1]) or np.isinf(attributes[i][-1]):
attributes[i][-1] = 0.0
# Add bonds
for bond in rdmol.GetBonds():
edge = Edge()
edge.i = bond.GetIdx()
edge.attributes = bondAttributes(bond)
edge.orderAtt = list(oneHotVector(bond.GetBondTypeAsDouble(), [1.0, 1.5, 2.0, 3.0]))
edge.aromAtt = list(oneHotVector(bond.GetIsAromatic(), [1.0, 0.0]))
edge.conjAtt = list(oneHotVector(bond.GetIsConjugated(), [1.0, 0.0]))
edge.ringAtt = list(oneHotVector(bond.IsInRing(), [1.0, 0.0]))
BeginAtom, EndAtom = bond.GetBeginAtom(), bond.GetEndAtom()
begin_idx, end_idx = BeginAtom.GetAtomicNum(), EndAtom.GetAtomicNum()
if begin_idx < end_idx:
bond_type = str(begin_idx) + '_' + str(end_idx)
else:
bond_type= str(end_idx) + '_' + str(begin_idx)
bond_attributes = []
bond_attributes = bond_attributes + list(oneHotVector(bond_type, bondtype_list_order))
edge.attributesAtt = np.array(bond_attributes, dtype=att_dtype)
edge.connects = (bond.GetBeginAtomIdx(), bond.GetEndAtomIdx())
graph.edges.append(edge)
# Add atoms
for k, atom in enumerate(rdmol.GetAtoms()):
node = Node()
node.i = atom.GetIdx()
node.attributes = atomAttributes(atom, extra_attributes = attributes[k])
node_type = atom.GetAtomicNum()
node_attributesAtt = []
node_attributesAtt = node_attributesAtt + list(oneHotVector(node_type, atomtype_list_order))
node.attributesAtt = np.array(node_attributesAtt, dtype=att_dtype)
for neighbor in atom.GetNeighbors():
node.neighbors.append((
neighbor.GetIdx(),
rdmol.GetBondBetweenAtoms(
atom.GetIdx(),
neighbor.GetIdx()
).GetIdx()
))
graph.nodes.append(node)
# Add counts, for convenience
graph.num_edges = len(graph.edges)
graph.num_nodes = len(graph.nodes)
return graph
def atom_level_descriptors(mol, include = ['functional'], asOneHot = False, ORIGINAL_VERSION = False):
"""
Given an RDKit mol, returns an N_atom-long list of lists,
each of which contains atom-level descriptors and their names
Returns:
(labels, attributes)
"""
attributes = [[] for i in mol.GetAtoms()]
labels = []
if 'functional' in include:
[attributes[i].append(x[0]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(mol))]
labels.append('Crippen contribution to logp')
[attributes[i].append(x[1]) \
for (i, x) in enumerate(rdMolDescriptors._CalcCrippenContribs(mol))]
labels.append('Crippen contribution to mr')
[attributes[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcTPSAContribs(mol))]
labels.append('TPSA contribution')
[attributes[i].append(x) \
for (i, x) in enumerate(rdMolDescriptors._CalcLabuteASAContribs(mol)[0])]
labels.append('Labute ASA contribution')
[attributes[i].append(x) \
for (i, x) in enumerate(EState.EStateIndices(mol))]
labels.append('EState Index')
rdPartialCharges.ComputeGasteigerCharges(mol)
[attributes[i].append(float(a.GetProp('_GasteigerCharge'))) \
for (i, a) in enumerate(mol.GetAtoms())]
labels.append('Gasteiger partial charge')
# Gasteiger partial charges sometimes gives NaN
for i in range(len(attributes)):
if np.isnan(attributes[i][-1]):
attributes[i][-1] = 0.0
[attributes[i].append(float(a.GetProp('_GasteigerHCharge'))) \
for (i, a) in enumerate(mol.GetAtoms())]
labels.append('Gasteiger hydrogen partial charge')
# Gasteiger partial charges sometimes gives NaN
for i in range(len(attributes)):
if np.isnan(attributes[i][-1]):
attributes[i][-1] = 0.0
if 'structural' in include:
[attributes[i].extend(atom_structural(mol.GetAtomWithIdx(i), asOneHot = asOneHot, ORIGINAL_VERSION = ORIGINAL_VERSION)) \
for i in range(len(attributes))]
labels.append('--many structural--')
if 'dftb' in include:
try:
dftb_atom_atts = atom_dftb(mol)
except ValueError as e:# often, an invalid element
print(e)
dftb_atom_atts = [[0 for i in range(18)] for j in range(mol.GetNumAtoms())]
except KeyError as e:
print(e)
dftb_atom_atts = [[0 for i in range(18)] for j in range(mol.GetNumAtoms())]
[attributes[i].extend(dftb_atom_atts[i]) for i in range(mol.GetNumAtoms())]
labels.append('--many DFTB--')
return (labels, attributes)
# lis2 = df['fps']
d2f = d2f.reset_index(drop=True)
lis = d2f['mol_blocks']
d2f.head(100)
mol_list = []
for m in lis:
m1 = Chem.MolFromMolBlock(m, removeHs=False)
mol_list.append(m1)
cdk2mol = [m for m in mol_list]
cdk2mol2 = copy.deepcopy(cdk2mol)
crippen_contribs = [
rdMolDescriptors._CalcCrippenContribs(mol) for mol in cdk2mol2
]
ref = cdk2mol_reference
ref_contrib = rdMolDescriptors._CalcCrippenContribs(ref)
targets = cdk2mol2[0:]
targets_contrib = crippen_contribs[0:]
for i, target in enumerate(targets):
crippenO3A = rdMolAlign.GetCrippenO3A(target, ref, targets_contrib[i],
ref_contrib)
crippenO3A.Align()
v.DeleteAll()
v.ShowMol(ref, name='ref', showOnly=False)
for i in range(len(targets)):
name = f'probe_{i}'
