def _sort_by_wiberg(self):
"""
Sorts the molecules by total Wiberg bond order around the trivalent
nitrogen.
"""
logging.info("Sorting molecules by Wiberg bond order")
self._mols.sort(key=lambda m: danceprops.get_dance_property(
m, self._properties).tri_n_bond_order)
def _write_bonds_to_csv(self):
"""Writes data about bonds around trivalent nitrogens to a CSV"""
logging.info(f"Writing bond data to {self._output_tri_n_bonds}")
with open(self._output_tri_n_bonds, 'w') as bondcsv:
bondcsv.write("bond_order,bond_length,element\n")
for mol in self._mols:
prop = danceprops.get_dance_property(mol, self._properties)
for bond in prop.tri_n_bonds:
bondcsv.write(f"{bond.bond_order},"
f"{bond.bond_length},"
f"{bond.element}\n")
def _write_data_to_csv(self):
"""Writes the data about trivalent nitrogens to a CSV"""
logging.info(f"Writing tri-n data to {self._output_tri_n_data}")
with open(self._output_tri_n_data, 'w') as datacsv:
datacsv.write(
"tri_n_bond_order,tri_n_bond_angle,tri_n_bond_length\n")
for mol in self._mols:
prop = danceprops.get_dance_property(mol, self._properties)
datacsv.write(f"{prop.tri_n_bond_order},"
f"{prop.tri_n_bond_angle},"
f"{prop.tri_n_bond_length}\n")
def test_append_properties_list_appends_small_list(
self, small_mols_and_props_list):
old_mols, old_props = small_mols_and_props_list
mols, props = old_mols.copy(), old_props.copy()
mols2 = [oechem.OEMol() for i in range(2)]
oechem.OESmilesToMol(mols2[0], "CN=C=O")
oechem.OESmilesToMol(mols2[1], "[C-]#[O+]")
danceprops.set_dance_property(mols2[0], 1)
danceprops.set_dance_property(mols2[1], 0)
old_mols2 = [oechem.OEMol(m) for m in mols2]
props2 = [danceprops.DanceProperties() for i in range(2)]
danceprops.append_properties_list(mols, props, mols2, props2)
# The properties obtained from the molecules put in the new list should
# be the same as the properties in the old list.
assert danceprops.get_dance_property(
mols[4],
props) == danceprops.get_dance_property(old_mols2[0], props2)
assert danceprops.get_dance_property(
mols[5],
props) == danceprops.get_dance_property(old_mols2[1], props2)
def assert_props_equal(self, mols: [oechem.OEMol], mol_indices: [int],
props: [danceprops.DanceProperties],
correct_props: [danceprops.DanceProperties],
correct_prop_indices: [int]):
"""
Assert that, for each molecule index given, the property returned with
danceprops.get_dance_property is equal to the property at the given
index in correct_props.
"""
if len(mol_indices) != len(correct_prop_indices):
raise RuntimeError("There must be an equal number of molecule and "
"correct property indices")
for i in range(len(mol_indices)):
assert danceprops.get_dance_property(
mols[mol_indices[i]],
props) == correct_props[correct_prop_indices[i]]
def write_mol_to_fingerprint_file(
mol: oechem.OEMol,
properties: [danceprops.DanceProperties],
select_output_dir: str,
select_bin_size: float,
wiberg_precision: float,
):
"""Writes a molecule to its appropriate SMILES fingerprint file"""
# Some of the molecules coming in may be invalid. DanceGenerator may find
# there was an error in charge calculations, in which case the charged
# copy was not assigned to the molecule. This function checks for that.
is_valid_molecule = \
lambda mol: mol.HasData(danceprops.DANCE_CHARGED_COPY_KEY)
if not is_valid_molecule(mol):
logging.debug(f"Ignored molecule {mol.GetTitle()}")
return
charged_copy = mol.GetData(danceprops.DANCE_CHARGED_COPY_KEY)
for atom in charged_copy.GetAtoms(oechem.OEIsInvertibleNitrogen()):
tri_n = atom
break
fingerprint = danceprops.DanceFingerprint(tri_n, wiberg_precision)
# Retrieve the total bond order around the trivalent nitrogen
bond_order = danceprops.get_dance_property(mol, properties).tri_n_bond_order
# Round the total bond order down to the lowest multiple of bin_size. For
# instance, if bin_size is 0.02, and the bond_order is 2.028, it becomes
# 2.02. This works because (bond_order / self._bin_size) generates a
# multiple of the bin_size. Then floor() finds the next integer less than
# the multiple. Finally, multiplying back by bin_size obtains the nearest
# actual value.
bond_order = math.floor(bond_order / select_bin_size) * select_bin_size
filename = f"{select_output_dir}/{bond_order},{fingerprint}.smi"
with open(filename, "a") as f:
f.write(f"{oechem.OEMolToSmiles(mol)} {mol.GetTitle()}\n")
logging.debug(f"Wrote {mol.GetTitle()} to {filename}")