def from_rdkit(cls, rd_atom: Chem.Atom) -> "Atom":
"""
Build a QUBEKit atom from an rdkit atom instance.
"""
atomic_number = rd_atom.GetAtomicNum()
index = rd_atom.GetIdx()
formal_charge = rd_atom.GetFormalCharge()
aromatic = rd_atom.GetIsAromatic()
bonds = [a.GetIdx() for a in rd_atom.GetNeighbors()]
# check for names in the normal places pdb, mol2 and mol
if rd_atom.HasProp("_Name"):
name = rd_atom.GetProp("_Name")
elif rd_atom.HasProp("_TriposAtomName"):
name = rd_atom.GetProp("_TriposAtomName")
else:
try:
name = rd_atom.GetMonomerInfo().GetName().strip()
except AttributeError:
name = None
# stereochem
if rd_atom.HasProp("_CIPCode"):
stereo_code = rd_atom.GetProp("_CIPCode")
else:
stereo_code = None
return cls(
atomic_number=atomic_number,
atom_index=index,
atom_name=name,
formal_charge=formal_charge,
aromatic=aromatic,
stereochemistry=stereo_code,
bonds=bonds,
)
def _get_atom_valence(self, atom: Chem.Atom):
"""
Cannot get the normal way as it cannot be sanitised.
:param atom:
:return:
"""
valence = 0
for bond in atom.GetBonds():
valence += bond.GetBondTypeAsDouble()
return valence - atom.GetFormalCharge()
def get_atom_type(self, atom: Chem.Atom) -> int:
"""
Get the atom type (represented as the index in self.atom_types) of the
given atom `atom`
Args:
atom (Chem.Atom):
The input atom
Returns:
int:
The atom type as int
"""
atom_symbol = atom.GetSymbol()
atom_charge = atom.GetFormalCharge()
atom_hs = atom.GetNumExplicitHs()
return self.atom_types.index((atom_symbol, atom_charge, atom_hs))
def atom_to_string(atom: Chem.Atom) -> str:
return '{}|{}|{}|{}|{}|{}|{}|{}|{}|{}|{}|{}|{}|{}'.format(
atom.GetChiralTag(),
atom.GetDegree(),
atom.GetExplicitValence(),
atom.GetFormalCharge(),
atom.GetHybridization(),
atom.GetIsAromatic(),
atom.GetMass(),
atom.GetNumExplicitHs(),
atom.GetNumImplicitHs(),
atom.GetNumRadicalElectrons(),
atom.GetTotalDegree(),
atom.GetTotalNumHs(),
atom.GetTotalValence(),
atom.IsInRing()
)
def atom_feature(atom: Atom) -> Tuple[torch.Tensor, int]:
'''
Extract atom feature vector.
Returns: (feature_vec, feature_dim)
Features:
# item # dim
atomic number 25
exp-valence 7
imp-valence 4
hybridization 7
Hs 5
degree 6
formal charge 6
in ring 1
aromatic 1
mass / 100 1
(sum) 63
'''
index = torch.tensor([
ATOM_MAP[atom.GetAtomicNum()],
atom.GetExplicitValence(),
atom.GetImplicitValence(),
HYBRID_MAP[atom.GetHybridization()],
atom.GetTotalNumHs(),
atom.GetDegree(),
atom.GetFormalCharge() + 2, # -2 ~ 3
int(atom.IsInRing()),
int(atom.GetIsAromatic())
]) + OFFSET
vec = torch.zeros(ATOM_FDIM)
vec[index] = 1.0
vec[-1] = atom.GetMass() / 100
return vec, ATOM_FDIM
def _parse_atom(self, atom: Chem.Atom) -> None:
self.log.debug(
f'Parsing {atom.GetSymbol()} at position {atom.GetIdx()}')
if atom.GetSymbol() == '*':
neighbor = atom.GetNeighbors()[0]
n_name = self._get_PDBInfo_atomname(neighbor)
if self.is_aminoacid() and neighbor.GetSymbol() == 'N':
# atom_name, index, connect_type, connect_name
self.CONNECT.append([n_name, 1, 'LOWER_CONNECT', 'LOWER'])
elif self.is_aminoacid() and neighbor.GetSymbol() == 'C':
# atom_name, index, connect_type, connect_name
self.CONNECT.append([n_name, 2, 'UPPER_CONNECT', 'UPPER'])
elif self.is_aminoacid():
i = max(3, len(self.CONNECT) + 1)
self.CONNECT.append([n_name, i, 'CONNECT'])
else:
self.CONNECT.append([n_name, len(self.CONNECT) + 1, 'CONNECT'])
else:
d = self._get_atom_descriptors(
atom) # dict of 'name', 'rtype': 'mtype', 'partial'
self.ATOM.append(d)
formal = atom.GetFormalCharge()
if formal != 0:
self.CHARGE.append([d['name'], formal])
def _copy_atom(atom: Chem.Atom) -> Chem.Atom:
new_atom = Chem.Atom(atom.GetSymbol())
new_atom.SetFormalCharge(atom.GetFormalCharge())
new_atom.SetNumRadicalElectrons(atom.GetNumRadicalElectrons())
return new_atom
def charge(atom: Atom) -> int:
"""Formal charge (int).
"""
return atom.GetFormalCharge()