def mset_to_smiles(max_atoms, filename):
mset = MSet()
try:
mset.load(filename)
except json.JSONDecodeError:
print("json decode error")
return (None, None)
if mset.n_atoms > max_atoms:
return (None, None)
geom = mset.get_min_geom
try:
charges = [
float(atom.labels['wB97X-D.6-311g**.charges'])
for atom in geom.atoms
]
except KeyError:
charges = [
float(atom.labels['wb97x-d.6-311gss.mulliken_charges'])
for atom in geom.atoms
]
charge = sum(charges) / len(charges)
try:
mol = xyz2mol(geom.at_nums, [atom.xyz for atom in geom.atoms], charge)
except Chem.AtomValenceException:
print("atom valence exception")
return (None, None)
smiles = Chem.MolToSmiles(mol, isomericSmiles=False)
m = Chem.MolFromSmiles(smiles)
smiles = Chem.MolToSmiles(m, isomericSmiles=False)
mset.identifiers.update({"smiles": smiles})
mset.save(filename)
return (filename, smiles)
def mset_to_smiles(filename):
mset = MSet()
try:
mset.load(filename)
except json.JSONDecodeError:
print("json decode error")
return None
geom = get_min_geom(mset.geometries)
charge = 0
try:
smiles = geom_to_smiles(geom, charge)
except:
return None
mset.identifiers.update({"smiles": smiles})
mset.save(filename)
return filename
def mset_to_smiles(filename):
mset = MSet()
mset.load(filename)
geom = mset.get_min_geom
try:
cm5_charges = [
float(atom.labels['wb97x_dz.cm5_charges']) for atom in geom.atoms
]
hirshfeld_charges = [
float(atom.labels['wb97x_dz.hirshfeld_charges'])
for atom in geom.atoms
]
charge = ((sum(cm5_charges) / len(cm5_charges)) +
(sum(hirshfeld_charges) / len(hirshfeld_charges))) / 2
except:
charge = 0
try:
smiles = geom_to_smiles(geom, charge)
except:
return None
mset.identifiers.update({"smiles": smiles})
mset.save(filename)
return filename
def load_ani1x(path_to_h5file, data_keys=[]):
# Example for extracting DFT/DZ energies and forces
for i, data in enumerate(iter_data_buckets(path_to_h5file,
keys=data_keys)):
atoms = [Atom(at_num) for at_num in data['atomic_numbers'].tolist()]
mset = MoleculeSet(atoms)
mset.filename = "/mnt/sdb1/adriscoll/ani1x-data/ani1x-msets/ani1x-mol" + str(
i) + ".mset"
mol_keys, atom_keys, geoms = [], [], []
for key in data.keys():
if key == 'atomic_numbers' or key == 'coordinates':
continue
elif 'energy' in key or 'dipole' in key:
mol_keys.append(key)
elif 'force' in key or 'charge' in key:
atom_keys.append(key)
mol_labels = {key: data[key][-1].tolist() for key in mol_keys}
atom_labels = {key: data[key][-1].tolist() for key in atom_keys}
geoms.append(
mset.new_geometry(data['coordinates'][-1].tolist(), mol_labels,
atom_labels))
mset.trajectories['ani.data'] = geoms
mset.save()
return
opt_natoms[len(opt_mol_data['atoms'])].append(opt_mol)
else:
opt_natoms[len(opt_mol_data['atoms'])] = [opt_mol]
with open(
"/mnt/sdb1/jeherr/chemspider_data/chno_msets/chno_opt_natoms.txt",
"w") as f:
json.dump(opt_natoms, f)
opt_matches = {}
meta_matches = {}
for n_atoms, meta_mols in meta_natoms.items():
opt_mols = opt_natoms[n_atoms]
opt_msets = []
meta_msets = []
for opt_mol in opt_mols:
opt_mset = MoleculeSet()
opt_mset.load(opt_mol)
opt_mset.filename = opt_mol
opt_msets.append(opt_mset)
for meta_mol in meta_mols:
meta_mset = MoleculeSet()
meta_mset.load(meta_mol)
meta_mset.filename = meta_mol
meta_msets.append(meta_mset)
for meta_mset in meta_msets:
matches = []
for opt_mset in opt_msets:
if meta_mset.compare_hash(opt_mset):
matches.append(opt_mset.filename)
meta_matches[meta_mset.filename] = matches
for opt_mset in opt_msets:
else:
opt_natoms[len(opt_mol_data['atoms'])] = [opt_mol]
with open(
"/mnt/sdb1/adriscoll/chemspider_data/expanded_msets/opt_smiles_natoms.txt",
"w") as file:
json.dump(opt_natoms, file)
opt_matches = {}
meta_matches = {}
for n_atoms, meta_mols in meta_natoms.items():
if n_atoms in opt_natoms.keys():
opt_mols = opt_natoms[n_atoms]
opt_msets = []
meta_msets = []
for opt_mol in opt_mols:
opt_mset = MoleculeSet()
opt_mset.load(opt_mol)
opt_mset.filename = opt_mol
opt_msets.append(opt_mset)
for meta_mol in meta_mols:
meta_mset = MoleculeSet()
meta_mset.load(meta_mol)
meta_mset.filename = meta_mol
meta_msets.append(meta_mset)
for meta_mset in meta_msets:
matches = []
for opt_mset in opt_msets:
if meta_mset.identifiers['smiles'] == opt_mset.identifiers[
'smiles']:
matches.append(opt_mset.filename)
meta_matches[meta_mset.filename] = matches
def read_opt_data(filename):
n_atoms = None
mset = None
atomic_nums = []
coords = []
energies = []
forces = []
dipoles = []
quadrupoles = []
charges = []
print("Reading ", filename)
with open(filename, "r") as f:
try:
while True:
line = next(f)
if "User input:" in line:
n_atoms, at_sym, method, basis = parse_sp_input(f)
elif "Standard Nuclear Orientation" in line:
atom_nums, coord = parse_atoms_coords(f, n_atoms)
if atom_nums is None:
return None
atomic_nums.append(atom_nums)
coords.append(coord)
# energy
elif "Convergence failure" in line:
return None
elif "Cycle" in line and "Energy" in line:
energies.append(parse_energy(f))
# forces
elif "Gradient of SCF Energy" in line:
force = parse_forces(f, n_atoms)
if force is not None:
forces.append(force)
else:
print(filename, " contains unparsed forces")
# dipoles
elif "Dipole Moment (Debye)" in line:
dipoles.append(parse_dipole(f))
# quadrupoles
elif "Quadrupole Moments (Debye-Ang)" in line:
quadrupoles.append(parse_quadrupole(f))
# charges
elif "Ground-State Mulliken Net Atomic Charges" in line:
charges.append(parse_charges(f, n_atoms))
elif "Optimization Cycle" in line:
if len(atomic_nums) == len(energies) == len(forces) == len(
coords) == len(dipoles) == len(quadrupoles) == len(
charges):
if line.split()[-1] == "1":
try:
atoms = [
Atom(at_num) for at_num in atomic_nums[0]
]
mset = MoleculeSet(atoms)
opt_trajectory = []
except:
print("Error making MSet for ", filename)
return None
if len(energies) > len(opt_trajectory):
opt_trajectory.append(
build_new_geom(atomic_nums, coords, energies,
forces, dipoles, quadrupoles,
charges, method, basis))
elif "OPTIMIZATION CONVERGED" in line:
mset.trajectories[".".join(
(method, basis, "opt"))] = opt_trajectory
return mset
except StopIteration:
print("Hit EOF on ", filename)
try:
mset.trajectories[".".join(
(method, basis, "opt"))] = opt_trajectory
return mset
except UnboundLocalError:
print("No MSet built for ", filename)
return None
def read_multi_sp_data(filenames):
n_atoms = None
mset = None
atomic_nums = []
coords = []
energies = []
forces = []
dipoles = []
quadrupoles = []
charges = []
for filename in filenames:
with open(filename, "r") as f:
try:
while True:
line = next(f)
if "User input:" in line:
n_atoms, at_sym, method, basis = parse_sp_input(f)
elif "Standard Nuclear Orientation" in line:
atom_nums, coord = parse_atoms_coords(f, n_atoms)
atomic_nums.append(atom_nums)
coords.append(coord)
# energy
elif "Cycle" in line and "Energy" in line:
energies.append(parse_energy(f))
# forces
elif "Gradient of SCF Energy" in line:
force = parse_forces(f, n_atoms)
if force is not None:
forces.append(force)
else:
print(filename, " contains unparsed forces")
# dipoles
elif "Dipole Moment (Debye)" in line:
dipoles.append(parse_dipole(f))
# quadrupoles
elif "Quadrupole Moments (Debye-Ang)" in line:
quadrupoles.append(parse_quadrupole(f))
# charges
elif "Ground-State Mulliken Net Atomic Charges" in line:
charges.append(parse_charges(f, n_atoms))
elif "Thank you very much for using Q-Chem." in line:
if mset is None:
print("Attempting to make initial MSet for ",
filename)
try:
atoms = [
Atom(at_num) for at_num in atomic_nums[0]
]
mset = MoleculeSet(atoms)
meta_trajectory = []
except:
print("Error making MSet for ", filename)
return None
if len(atomic_nums) == len(energies) == len(
forces) == len(coords) == len(dipoles) == len(
quadrupoles) == len(charges):
if len(energies) > len(meta_trajectory):
meta_trajectory.append(
build_new_geom(atomic_nums, coords,
energies, forces, dipoles,
quadrupoles, charges,
method, basis))
else:
min_len = min([
len(atomic_nums),
len(energies),
len(forces),
len(coords),
len(dipoles),
len(quadrupoles),
len(charges)
])
atomic_nums, energies, forces, coords, dipoles, quadrupoles, charges = atomic_nums[
:min_len], \
energies[
:min_len], forces[
:min_len], coords[
:min_len], dipoles[
:min_len], quadrupoles[
:min_len], charges[
:min_len]
except (StopIteration, UnicodeDecodeError):
if mset is not None:
if len(atomic_nums) == len(energies) == len(forces) == len(
coords) == len(dipoles) == len(quadrupoles) == len(
charges):
if len(energies) > len(meta_trajectory):
meta_trajectory.append(
build_new_geom(atomic_nums, coords, energies,
forces, dipoles, quadrupoles,
charges, method, basis))
else:
min_len = min([
len(atomic_nums),
len(energies),
len(forces),
len(coords),
len(dipoles),
len(quadrupoles),
len(charges)
])
atomic_nums, energies, forces, coords, dipoles, quadrupoles, charges = atomic_nums[:min_len], \
energies[
:min_len], forces[
:min_len], coords[
:min_len], dipoles[
:min_len], quadrupoles[
:min_len], charges[
:min_len]
continue
if mset is None:
print("No geometries collected for ", filenames[0])
return None
mset.trajectories[".".join((method, basis, "meta"))] = meta_trajectory
return mset
def read_aimd_data(filename):
atomic_nums = []
coords = []
energies = []
forces = []
dipoles = []
quadrupoles = []
charges = []
print("Reading ", filename)
with open(filename, "r") as f:
try:
while True:
line = next(f)
if "User input:" in line:
n_atoms, at_sym, method, basis, time_steps, aimd_steps, aimd_temp = parse_aimd_input(
f)
# These next sections **SHOULD** only grab the first instance of these properties (i.e. before the AIMD
# section starts)
elif "Standard Nuclear Orientation" in line:
atom_nums, coord = parse_atoms_coords(f, n_atoms)
atomic_nums.append(atom_nums)
coords.append(coord)
# energy
elif "Cycle" in line and "Energy" in line:
energies.append(parse_energy(f))
# forces
elif "Gradient of SCF Energy" in line:
forces.append(parse_forces(f, n_atoms))
# dipoles
elif "Dipole Moment (Debye)" in line:
dipoles.append(parse_dipole(f))
# quadrupoles
elif "Quadrupole Moments (Debye-Ang)" in line:
quadrupoles.append(parse_quadrupole(f))
# charges
elif "Ground-State Mulliken Net Atomic Charges" in line:
charges.append(parse_charges(f, n_atoms))
elif "AB INITIO MOLECULAR DYNAMICS" in line:
atoms = [Atom(at_num) for at_num in atomic_nums[0]]
mset = MoleculeSet(atoms)
aimd_trajectory = [
build_new_geom(atomic_nums, coords, energies, forces,
dipoles, quadrupoles, charges, method,
basis)
]
while True:
line = next(f)
if "TIME STEPS COMPLETED" in line:
if len(atomic_nums) == len(energies) == len(
forces) == len(coords) == len(
dipoles) == len(quadrupoles) == len(
charges):
if len(energies) > len(aimd_trajectory):
aimd_trajectory.append(
build_new_geom(atomic_nums, coords,
energies, forces,
dipoles, quadrupoles,
charges, method, basis))
else:
print(
"Error reading AIMD trajectory at last step",
filename,
"Returning MSet before this step")
mset.trajectories[".".join(
(method, basis, "aimd"))] = aimd_trajectory
return mset
elif "TIME STEP" in line:
time_step = int(line.split()[2].lstrip("#"))
time_au = float(line.split()[5])
time_fs = float(line.split()[8])
if len(atomic_nums) == len(energies) == len(
forces) == len(coords) == len(
dipoles) == len(quadrupoles) == len(
charges):
aimd_trajectory.append(
build_new_geom(atomic_nums, coords,
energies, forces, dipoles,
quadrupoles, charges,
method, basis))
else:
print("Error reading AIMD trajectory at step",
time_step, filename,
"Returning MSet before "
"this step")
mset.trajectories[".".join(
(method, basis, "aimd"))] = aimd_trajectory
return mset
elif "Standard Nuclear Orientation" in line:
atom_nums, coord = parse_atoms_coords(f, n_atoms)
atomic_nums.append(atom_nums)
coords.append(coord)
# energy
elif "Cycle" in line and "Energy" in line:
energies.append(parse_energy(f))
# forces
elif "Gradient of SCF Energy" in line:
forces.append(parse_forces(f, n_atoms))
# dipoles
elif "Dipole Moment (Debye)" in line:
dipoles.append(parse_dipole(f))
# quadrupoles
elif "Quadrupole Moments (Debye-Ang)" in line:
quadrupoles.append(parse_quadrupole(f))
# charges
elif "Ground-State Mulliken Net Atomic Charges" in line:
charges.append(parse_charges(f, n_atoms))
except StopIteration:
print("Hit EOF on ", filename)
try:
mset.trajectories[".".join(
(method, basis, "aimd"))] = aimd_trajectory
return mset
except UnboundLocalError:
print("No MSet built for ", filename)
return None