def _mol_from_rdkit(self):
"""
Using an RDKit Molecule object, extract the name, topology, coordinates and atoms
"""
if self.name is None:
self.name = self.rdkit_mol.GetProp("_Name")
atoms = []
self.topology = nx.Graph()
# Collect the atom names and bonds
for atom in self.rdkit_mol.GetAtoms():
# Collect info about each atom
atomic_number = atom.GetAtomicNum()
index = atom.GetIdx()
try:
# PDB file extraction
atom_name = atom.GetMonomerInfo().GetName().strip()
except AttributeError:
try:
# Mol2 file extraction
atom_name = atom.GetProp("_TriposAtomName")
except KeyError:
# smiles and mol files have no atom names so generate them here if they are not declared
atom_name = f"{atom.GetSymbol()}{index}"
qube_atom = Atom(atomic_number,
index,
atom_name,
formal_charge=atom.GetFormalCharge())
# Instance the basic qube_atom
qube_atom.atom_type = atom.GetSmarts()
# Add the atoms as nodes
self.topology.add_node(atom.GetIdx())
# Add the bonds
for bonded in atom.GetNeighbors():
self.topology.add_edge(atom.GetIdx(), bonded.GetIdx())
qube_atom.add_bond(bonded.GetIdx())
# Now add the atom to the molecule
atoms.append(qube_atom)
self.coords = self.rdkit_mol.GetConformer().GetPositions()
self.atoms = atoms or None
def _read_mol2(self):
"""
Internal mol2 reader. Only called when RDKit failed to read the mol2.
Extracts the topology, atoms and coords of the molecule.
"""
coords = []
self.topology = nx.Graph()
atoms = []
atom_count = 0
with open(self.mol_input, 'r') as mol2:
atom_flag = False
bond_flag = False
for line in mol2:
if '@<TRIPOS>ATOM' in line:
atom_flag = True
continue
elif '@<TRIPOS>BOND' in line:
atom_flag = False
bond_flag = True
continue
elif '@<TRIPOS>SUBSTRUCTURE' in line:
bond_flag = False
continue
if atom_flag:
# Add the molecule information
atomic_symbol = line.split()[1][:2]
atomic_symbol = re.sub('[0-9]+', '', atomic_symbol)
atomic_symbol = atomic_symbol.strip().title()
atomic_number = Element().number(atomic_symbol)
coords.append([float(line.split()[2]), float(line.split()[3]), float(line.split()[4])])
# Collect the atom names
atom_name = str(line.split()[1])
# Add the nodes to the topology object
self.topology.add_node(atom_count)
atom_count += 1
# Get the atom types
atom_type = line.split()[5]
atom_type = atom_type.replace(".", "")
# Make the qube_atom
qube_atom = Atom(atomic_number, atom_count, atom_name)
qube_atom.atom_type = atom_type
atoms.append(qube_atom)
if bond_flag:
# Add edges to the topology network
atom_index, bonded_index = int(line.split()[1]) - 1, int(line.split()[2]) - 1
self.topology.add_edge(atom_index, bonded_index)
atoms[atom_index].add_bond(bonded_index)
atoms[bonded_index].add_bond(atom_index)
# put the object back into the correct place
self.coords = np.array(coords)
self.atoms = atoms or None