def test_truncate_n_molecules():
substance = Substance()
substance.add_component(
component=Component(smiles="[Na+]"),
amount=MoleFraction(0.00267),
)
substance.add_component(
component=Component(smiles="[Cl-]"),
amount=MoleFraction(0.00267),
)
substance.add_component(component=Component(smiles="O"),
amount=MoleFraction(1.0 - 2.0 * 0.00267))
# Attempt to get the number of molecules without truncating.
with pytest.raises(ValueError):
substance.get_molecules_per_component(1000, truncate_n_molecules=False)
# Attempt to get the number of molecules with truncating.
molecule_counts = substance.get_molecules_per_component(
1000, truncate_n_molecules=True)
assert molecule_counts == {
"[Na+]{solv}": 3,
"[Cl-]{solv}": 3,
"O{solv}": 994
}
def test_solvate_existing_structure_protocol():
"""Tests solvating a single methanol molecule in water."""
import mdtraj
methanol_component = Component("CO")
methanol_substance = Substance()
methanol_substance.add_component(methanol_component, ExactAmount(1))
water_substance = Substance()
water_substance.add_component(Component("O"), MoleFraction(1.0))
with tempfile.TemporaryDirectory() as temporary_directory:
build_methanol_coordinates = BuildCoordinatesPackmol("build_methanol")
build_methanol_coordinates.max_molecules = 1
build_methanol_coordinates.substance = methanol_substance
build_methanol_coordinates.execute(temporary_directory,
ComputeResources())
methanol_residue_name = build_methanol_coordinates.assigned_residue_names[
methanol_component.identifier]
solvate_coordinates = SolvateExistingStructure("solvate_methanol")
solvate_coordinates.max_molecules = 9
solvate_coordinates.substance = water_substance
solvate_coordinates.solute_coordinate_file = (
build_methanol_coordinates.coordinate_file_path)
solvate_coordinates.execute(temporary_directory, ComputeResources())
solvated_system = mdtraj.load_pdb(
solvate_coordinates.coordinate_file_path)
assert solvated_system.n_residues == 10
assert solvated_system.top.residue(0).name == methanol_residue_name
def test_multiple_amounts():
substance = Substance()
sodium = Component("[Na+]")
chloride = Component("[Cl-]")
substance.add_component(sodium, MoleFraction(0.75))
substance.add_component(sodium, ExactAmount(1))
substance.add_component(chloride, MoleFraction(0.25))
substance.add_component(chloride, ExactAmount(1))
assert substance.number_of_components == 2
sodium_amounts = substance.get_amounts(sodium)
chlorine_amounts = substance.get_amounts(chloride)
assert len(sodium_amounts) == 2
assert len(chlorine_amounts) == 2
molecule_counts = substance.get_molecules_per_component(6)
assert len(molecule_counts) == 2
assert molecule_counts[sodium.identifier] == 4
assert molecule_counts[chloride.identifier] == 2
def dummy_complex() -> Substance:
substance = Substance()
substance.add_component(
Component(smiles="C", role=Component.Role.Ligand), ExactAmount(1)
)
substance.add_component(
Component(smiles="CO", role=Component.Role.Receptor), ExactAmount(1)
)
return substance
def create_dummy_substance(number_of_components, elements=None):
"""Creates a substance with a given number of components,
each containing the specified elements.
Parameters
----------
number_of_components : int
The number of components to add to the substance.
elements : list of str
The elements that each component should containt.
Returns
-------
Substance
The created substance.
"""
if elements is None:
elements = ["C"]
substance = Substance()
mole_fraction = 1.0 / number_of_components
for index in range(number_of_components):
smiles_pattern = "".join(elements * (index + 1))
substance.add_component(Component(smiles_pattern),
MoleFraction(mole_fraction))
return substance
def from_components(cls, *components):
"""Creates a new `Substance` object from a list of components.
This method assumes that all components should be present with
equal mole fractions.
Parameters
----------
components: Component or str
The components to add to the substance. These may either be full
`Component` objects or just the smiles representation
of the component.
Returns
-------
Substance
The substance containing the requested components in equal amounts.
"""
if len(components) == 0:
raise ValueError("At least one component must be specified")
mole_fraction = 1.0 / len(components)
return_substance = cls()
for component in components:
if isinstance(component, str):
component = Component(smiles=component)
return_substance.add_component(component,
MoleFraction(mole_fraction))
return return_substance
def test_build_docked_coordinates_protocol():
"""Tests docking a methanol molecule into alpha-Cyclodextrin."""
if not has_openeye():
pytest.skip("The `BuildDockedCoordinates` protocol requires OpenEye.")
ligand_substance = Substance()
ligand_substance.add_component(
Component("CO", role=Component.Role.Ligand),
ExactAmount(1),
)
# TODO: This test could likely be made substantially faster
# by storing the binary prepared receptor. Would this
# be in breach of any oe license terms?
with tempfile.TemporaryDirectory() as temporary_directory:
build_docked_coordinates = BuildDockedCoordinates("build_methanol")
build_docked_coordinates.ligand_substance = ligand_substance
build_docked_coordinates.number_of_ligand_conformers = 5
build_docked_coordinates.receptor_coordinate_file = get_data_filename(
"test/molecules/acd.mol2")
build_docked_coordinates.execute(temporary_directory,
ComputeResources())
docked_pdb = PDBFile(
build_docked_coordinates.docked_complex_coordinate_path)
assert docked_pdb.topology.getNumResidues() == 2
def test_add_mole_fractions():
substance = Substance()
substance.add_component(Component("C"), MoleFraction(0.5))
substance.add_component(Component("C"), MoleFraction(0.5))
assert substance.number_of_components == 1
amounts = substance.get_amounts(substance.components[0])
assert len(amounts) == 1
amount = next(iter(amounts))
assert isinstance(amount, MoleFraction)
assert np.isclose(amount.value, 1.0)
def test_build_coordinates_packmol_exact(count_exact_amount):
"""Tests that the build coordinate protocol behaves correctly for substances
with exact amounts."""
import mdtraj
substance = Substance()
substance.add_component(Component("O"), MoleFraction(1.0))
substance.add_component(Component("C"), ExactAmount(1))
max_molecule = 11 if count_exact_amount else 10
build_coordinates = BuildCoordinatesPackmol("build_coordinates")
build_coordinates.max_molecules = max_molecule
build_coordinates.count_exact_amount = count_exact_amount
build_coordinates.substance = substance
with tempfile.TemporaryDirectory() as directory:
build_coordinates.execute(directory)
built_system = mdtraj.load_pdb(build_coordinates.coordinate_file_path)
assert built_system.n_residues == 11
def data_frame() -> pandas.DataFrame:
temperatures = [298.15, 318.15]
pressures = [101.325, 101.0]
properties = [Density, EnthalpyOfMixing]
mole_fractions = [(1.0, ), (1.0, ), (0.25, 0.75), (0.75, 0.25)]
smiles = {1: [("C(F)(Cl)(Br)", ), ("C", )], 2: [("CO", "C"), ("C", "CO")]}
loop_variables = [(
temperature,
pressure,
property_type,
mole_fraction,
) for temperature in temperatures for pressure in pressures
for property_type in properties
for mole_fraction in mole_fractions]
data_entries = []
for temperature, pressure, property_type, mole_fraction in loop_variables:
n_components = len(mole_fraction)
for smiles_tuple in smiles[n_components]:
substance = Substance()
for smiles_pattern, x in zip(smiles_tuple, mole_fraction):
substance.add_component(Component(smiles_pattern),
MoleFraction(x))
data_entries.append(
property_type(
thermodynamic_state=ThermodynamicState(
temperature=temperature * unit.kelvin,
pressure=pressure * unit.kilopascal,
),
phase=PropertyPhase.Liquid,
value=1.0 * property_type.default_unit(),
uncertainty=1.0 * property_type.default_unit(),
source=MeasurementSource(doi=" "),
substance=substance,
))
data_set = PhysicalPropertyDataSet()
data_set.add_properties(*data_entries)
return data_set.to_pandas()
def create_substance():
test_substance = Substance()
test_substance.add_component(
Component("C", role=Component.Role.Solute),
ExactAmount(1),
)
test_substance.add_component(
Component("CC", role=Component.Role.Ligand),
ExactAmount(1),
)
test_substance.add_component(
Component("CCC", role=Component.Role.Receptor),
ExactAmount(1),
)
test_substance.add_component(
Component("O", role=Component.Role.Solvent),
MoleFraction(1.0),
)
return test_substance
def _build_input_output_substances():
"""Builds sets if input and expected substances for the
`test_build_coordinate_composition` test.
Returns
-------
list of tuple of Substance and Substance
A list of input and expected substances.
"""
# Start with some easy cases
substances = [
(Substance.from_components("O"), Substance.from_components("O")),
(Substance.from_components("O",
"C"), Substance.from_components("O", "C")),
(
Substance.from_components("O", "C", "CO"),
Substance.from_components("O", "C", "CO"),
),
]
# Handle some cases where rounding will need to occur.
input_substance = Substance()
input_substance.add_component(Component("O"), MoleFraction(0.41))
input_substance.add_component(Component("C"), MoleFraction(0.59))
expected_substance = Substance()
expected_substance.add_component(Component("O"), MoleFraction(0.4))
expected_substance.add_component(Component("C"), MoleFraction(0.6))
substances.append((input_substance, expected_substance))
input_substance = Substance()
input_substance.add_component(Component("O"), MoleFraction(0.59))
input_substance.add_component(Component("C"), MoleFraction(0.41))
expected_substance = Substance()
expected_substance.add_component(Component("O"), MoleFraction(0.6))
expected_substance.add_component(Component("C"), MoleFraction(0.4))
substances.append((input_substance, expected_substance))
return substances
def _build_substance(
guest_smiles: Optional[str],
host_smiles: str,
ionic_strength: Optional[unit.Quantity],
negative_buffer_ion: str = "[Cl-]",
positive_buffer_ion: str = "[Na+]",
):
"""Builds a substance containing a ligand and receptor solvated in an aqueous
solution with a given ionic strength
Parameters
----------
guest_smiles
The SMILES descriptor of the guest.
host_smiles
The SMILES descriptor of the host.
ionic_strength
The ionic strength of the aqueous solvent.
Returns
-------
The built substance.
"""
from openff.toolkit.topology import Molecule
from simtk import unit as simtk_unit
substance = Substance()
if guest_smiles is not None:
guest = Component(smiles=guest_smiles, role=Component.Role.Ligand)
substance.add_component(component=guest, amount=ExactAmount(1))
host = Component(smiles=host_smiles, role=Component.Role.Receptor)
substance.add_component(component=host, amount=ExactAmount(1))
water = Component(smiles="O", role=Component.Role.Solvent)
sodium = Component(smiles=positive_buffer_ion,
role=Component.Role.Solvent)
chlorine = Component(smiles=negative_buffer_ion,
role=Component.Role.Solvent)
water_mole_fraction = 1.0
if ionic_strength is not None:
salt_mole_fraction = Substance.calculate_aqueous_ionic_mole_fraction(
ionic_strength)
if isinstance(salt_mole_fraction, unit.Quantity):
# noinspection PyUnresolvedReferences
salt_mole_fraction = salt_mole_fraction.magnitude
water_mole_fraction = 1.0 - salt_mole_fraction * 2
substance.add_component(
component=sodium,
amount=MoleFraction(salt_mole_fraction),
)
substance.add_component(
component=chlorine,
amount=MoleFraction(salt_mole_fraction),
)
substance.add_component(component=water,
amount=MoleFraction(water_mole_fraction))
host_molecule_charge = Molecule.from_smiles(host_smiles).total_charge
guest_molecule_charge = (
0.0 * simtk_unit.elementary_charge if guest_smiles is None else
Molecule.from_smiles(guest_smiles).total_charge)
net_charge = (host_molecule_charge +
guest_molecule_charge).value_in_unit(
simtk_unit.elementary_charge)
n_counter_ions = abs(int(net_charge))
if net_charge <= -0.9999:
substance.add_component(sodium, ExactAmount(n_counter_ions))
elif net_charge >= 0.9999:
substance.add_component(chlorine, ExactAmount(n_counter_ions))
return substance
def _apply(
cls,
data_frame: pandas.DataFrame,
schema: ImportFreeSolvSchema,
n_processes,
) -> pandas.DataFrame:
from openff.evaluator import properties, substances, unit
# Convert the data frame into data rows.
free_solv_data_frame = cls._download_free_solv()
data_entries = []
for _, row in free_solv_data_frame.iterrows():
# Extract and standardize the SMILES pattern of the
solute_smiles = row["SMILES"].lstrip().rstrip()
solute_smiles = substances.Component(solute_smiles).smiles
# Build the substance.
substance = Substance()
substance.add_component(Component(smiles="O"), MoleFraction(1.0))
substance.add_component(
Component(smiles=solute_smiles, role=Component.Role.Solute),
ExactAmount(1),
)
# Extract the value and uncertainty
value = (float(row["experimental value (kcal/mol)"]) *
unit.kilocalorie / unit.mole)
std_error = (float(row["experimental uncertainty (kcal/mol)"]) *
unit.kilocalorie / unit.mole)
# Attempt to extract a DOI
original_source = row[
"experimental reference (original or paper this value was taken from)"]
doi = cls._validate_doi(original_source)
data_entry = SolvationFreeEnergy(
thermodynamic_state=ThermodynamicState(
temperature=298.15 * unit.kelvin,
pressure=101.325 * unit.kilopascal,
),
phase=PropertyPhase.Liquid,
substance=substance,
value=value.to(properties.SolvationFreeEnergy.default_unit()),
uncertainty=std_error.to(
properties.SolvationFreeEnergy.default_unit()),
source=MeasurementSource(doi=doi),
)
data_entries.append(data_entry)
data_set = PhysicalPropertyDataSet()
data_set.add_properties(*data_entries)
free_solv_data_frame = data_set.to_pandas()
data_frame = pandas.concat([data_frame, free_solv_data_frame],
ignore_index=True,
sort=False)
return data_frame
def main():
os.makedirs("raw_data_v2", exist_ok=True)
for data_set_name in [
"curated_data_set",
"gaff 1.81",
"gaff 2.11",
"parsley 1.0.0",
"smirnoff99frosst 1.1.0",
]:
with open(os.path.join("raw_data", f"{data_set_name}.json")) as file:
raw_data_set = json.load(file)
assert (raw_data_set["@type"] ==
"propertyestimator.datasets.datasets.PhysicalPropertyDataSet")
physical_properties = []
for raw_data_set_entries in raw_data_set["properties"].values():
for raw_data_set_entry in raw_data_set_entries:
# Extract the substance this entry was measured for.
substance = Substance()
for raw_component in raw_data_set_entry["substance"][
"components"]:
component = Component(
smiles=raw_component["smiles"],
role=Component.Role[raw_component["role"]["value"]],
)
raw_amounts = raw_data_set_entry["substance"]["amounts"][
raw_component["smiles"]]
for raw_amount in raw_amounts["value"]:
if (raw_amount["@type"] ==
"propertyestimator.substances.Substance->MoleFraction"
):
substance.add_component(
component, MoleFraction(raw_amount["value"]))
elif (raw_amount["@type"] ==
"propertyestimator.substances.Substance->ExactAmount"
):
substance.add_component(
component, ExactAmount(raw_amount["value"]))
else:
raise NotImplementedError()
# Extract the source of the property
if (raw_data_set_entry["source"]["@type"] ==
"propertyestimator.properties.properties.CalculationSource"
):
source = CalculationSource(
fidelity=raw_data_set_entry["source"]["fidelity"])
elif (raw_data_set_entry["source"]["@type"] ==
"propertyestimator.properties.properties.MeasurementSource"
):
source = MeasurementSource(doi=correct_doi(
raw_data_set_entry["source"]["reference"]))
else:
raise NotImplementedError()
# Generate the new property object.
property_class = getattr(
properties, raw_data_set_entry["@type"].split(".")[-1])
physical_property = property_class(
thermodynamic_state=ThermodynamicState(
temperature=(
raw_data_set_entry["thermodynamic_state"]
["temperature"]["value"] *
unit.Unit(raw_data_set_entry["thermodynamic_state"]
["temperature"]["unit"])),
pressure=(
raw_data_set_entry["thermodynamic_state"]
["pressure"]["value"] *
unit.Unit(raw_data_set_entry["thermodynamic_state"]
["pressure"]["unit"])),
),
phase=PropertyPhase(raw_data_set_entry["phase"]),
substance=substance,
value=(raw_data_set_entry["value"]["value"] *
unit.Unit(raw_data_set_entry["value"]["unit"])),
uncertainty=(
None if isinstance(source, MeasurementSource) else
(raw_data_set_entry["uncertainty"]["value"] *
unit.Unit(raw_data_set_entry["uncertainty"]["unit"])
)),
source=source,
)
physical_property.id = raw_data_set_entry["id"]
physical_properties.append(physical_property)
data_set = PhysicalPropertyDataSet()
data_set.add_properties(*physical_properties)
data_set.json(os.path.join("raw_data_v2", f"{data_set_name}.json"),
format=True)
data_set.to_pandas().to_csv(
os.path.join("raw_data_v2", f"{data_set_name}.csv"))
def test_solvation_yank_protocol(solvent_smiles):
full_substance = Substance()
full_substance.add_component(
Component(smiles="CO", role=Component.Role.Solute),
ExactAmount(1),
)
full_substance.add_component(
Component(smiles=solvent_smiles, role=Component.Role.Solvent),
MoleFraction(1.0),
)
solvent_substance = Substance()
solvent_substance.add_component(
Component(smiles=solvent_smiles, role=Component.Role.Solvent),
MoleFraction(1.0),
)
solute_substance = Substance()
solute_substance.add_component(
Component(smiles="CO", role=Component.Role.Solute),
ExactAmount(1),
)
thermodynamic_state = ThermodynamicState(temperature=298.15 * unit.kelvin,
pressure=1.0 * unit.atmosphere)
with tempfile.TemporaryDirectory() as directory:
with temporarily_change_directory(directory):
force_field_path = "ff.json"
with open(force_field_path, "w") as file:
file.write(build_tip3p_smirnoff_force_field().json())
solvated_coordinate_path, solvated_system = _setup_dummy_system(
"full", full_substance, 2, force_field_path)
vacuum_coordinate_path, vacuum_system = _setup_dummy_system(
"vacuum", solute_substance, 1, force_field_path)
run_yank = SolvationYankProtocol("yank")
run_yank.solute = solute_substance
run_yank.solvent_1 = solvent_substance
run_yank.solvent_2 = Substance()
run_yank.thermodynamic_state = thermodynamic_state
run_yank.number_of_iterations = 1
run_yank.steps_per_iteration = 1
run_yank.checkpoint_interval = 1
run_yank.verbose = True
run_yank.setup_only = True
run_yank.solution_1_coordinates = solvated_coordinate_path
run_yank.solution_1_system = solvated_system
run_yank.solution_2_coordinates = vacuum_coordinate_path
run_yank.solution_2_system = vacuum_system
run_yank.electrostatic_lambdas_1 = [1.00]
run_yank.steric_lambdas_1 = [1.00]
run_yank.electrostatic_lambdas_2 = [1.00]
run_yank.steric_lambdas_2 = [1.00]
run_yank.execute("", ComputeResources())
def test_ligand_receptor_yank_protocol():
full_substance = Substance()
full_substance.add_component(
Component(smiles="c1ccccc1", role=Component.Role.Receptor),
ExactAmount(1),
)
full_substance.add_component(
Component(smiles="C", role=Component.Role.Ligand),
ExactAmount(1),
)
full_substance.add_component(
Component(smiles="O", role=Component.Role.Solvent),
MoleFraction(1.0),
)
solute_substance = Substance()
solute_substance.add_component(
Component(smiles="C", role=Component.Role.Ligand),
ExactAmount(1),
)
solute_substance.add_component(
Component(smiles="O", role=Component.Role.Solvent),
MoleFraction(1.0),
)
thermodynamic_state = ThermodynamicState(temperature=298.15 * unit.kelvin,
pressure=1.0 * unit.atmosphere)
with tempfile.TemporaryDirectory() as directory:
with temporarily_change_directory(directory):
force_field_path = "ff.json"
with open(force_field_path, "w") as file:
file.write(build_tip3p_smirnoff_force_field().json())
complex_coordinate_path, complex_system = _setup_dummy_system(
"full", full_substance, 3, force_field_path)
ligand_coordinate_path, ligand_system = _setup_dummy_system(
"ligand", solute_substance, 2, force_field_path)
run_yank = LigandReceptorYankProtocol("yank")
run_yank.substance = full_substance
run_yank.thermodynamic_state = thermodynamic_state
run_yank.number_of_iterations = 1
run_yank.steps_per_iteration = 1
run_yank.checkpoint_interval = 1
run_yank.verbose = True
run_yank.setup_only = True
run_yank.ligand_residue_name = "TMP"
run_yank.receptor_residue_name = "TMP"
run_yank.solvated_ligand_coordinates = ligand_coordinate_path
run_yank.solvated_ligand_system = ligand_system
run_yank.solvated_complex_coordinates = complex_coordinate_path
run_yank.solvated_complex_system = complex_system
run_yank.force_field_path = force_field_path
run_yank.execute("", ComputeResources())
def _generate_residue_name(residue, smiles):
"""Generates residue name for a particular residue which
corresponds to a particular smiles pattern.
Where possible (i.e for amino acids and ions) a standard residue
name will be returned, otherwise a random name will be used.
Parameters
----------
residue: mdtraj.core.topology.Residue
The residue to assign the name to.
smiles: str
The SMILES pattern to generate a resiude name for.
"""
from mdtraj.core import residue_names
from openff.toolkit.topology import Molecule
# Define the set of residue names which should be discarded
# if randomly generated as they have a reserved meaning.
# noinspection PyProtectedMember
forbidden_residue_names = [
*residue_names._AMINO_ACID_CODES,
*residue_names._SOLVENT_TYPES,
*residue_names._WATER_RESIDUES,
"ADE",
"CYT",
"CYX",
"DAD",
"DGU",
"FOR",
"GUA",
"HID",
"HIE",
"HIH",
"HSD",
"HSH",
"HSP",
"NMA",
"THY",
"URA",
]
amino_residue_mappings = {
"C[C@H](N)C(=O)O": "ALA",
"N=C(N)NCCC[C@H](N)C(=O)O": "ARG",
"NC(=O)C[C@H](N)C(=O)O": "ASN",
"N[C@@H](CC(=O)O)C(=O)O": "ASP",
"N[C@@H](CS)C(=O)O": "CYS",
"N[C@@H](CCC(=O)O)C(=O)O": "GLU",
"NC(=O)CC[C@H](N)C(=O)O": "GLN",
"NCC(=O)O": "GLY",
"N[C@@H](Cc1c[nH]cn1)C(=O)O": "HIS",
"CC[C@H](C)[C@H](N)C(=O)O": "ILE",
"CC(C)C[C@H](N)C(=O)O": "LEU",
"NCCCC[C@H](N)C(=O)O": "LYS",
"CSCC[C@H](N)C(=O)O": "MET",
"N[C@@H](Cc1ccccc1)C(=O)O": "PHE",
"O=C(O)[C@@H]1CCCN1": "PRO",
"N[C@@H](CO)C(=O)O": "SER",
"C[C@@H](O)[C@H](N)C(=O)O": "THR",
"N[C@@H](Cc1c[nH]c2ccccc12)C(=O)O": "TRP",
"N[C@@H](Cc1ccc(O)cc1)C(=O)O": "TYR",
"CC(C)[C@H](N)C(=O)O": "VAL",
}
standardized_smiles = Component(smiles=smiles).smiles
# Check for amino acids.
if standardized_smiles in amino_residue_mappings:
residue.name = amino_residue_mappings[standardized_smiles]
return
# Check for water
if standardized_smiles == "O":
residue.name = "HOH"
# Re-assign the water atom names. These need to be set to get
# correct CONECT statements.
h_counter = 1
for atom in residue.atoms:
if atom.element.symbol == "O":
atom.name = "O1"
else:
atom.name = f"H{h_counter}"
h_counter += 1
return
# Check for ions
openff_molecule = Molecule.from_smiles(smiles, allow_undefined_stereo=True)
if openff_molecule.n_atoms == 1:
residue.name = _ion_residue_name(openff_molecule)
residue.atom(0).name = residue.name
return
# Randomly generate a name
random_residue_name = "".join(
[random.choice(string.ascii_uppercase) for _ in range(3)])
while random_residue_name in forbidden_residue_names:
# Re-choose the residue name until we find a safe one.
random_residue_name = "".join(
[random.choice(string.ascii_uppercase) for _ in range(3)])
residue.name = random_residue_name
# Assign unique atom names.
element_counter = defaultdict(int)
for atom in residue.atoms:
atom.name = f"{atom.element.symbol}{element_counter[atom.element.symbol] + 1}"
element_counter[atom.element.symbol] += 1
def complete_evaluator_data_set():
"""Create a more comprehensive `PhysicalPropertyDataSet` which contains one
measurement for each of:
* pure density
* binary density
* pure enthalpy of vaporization
* binary enthalpy of mixing
* binary excess molar volume
* hydration free energy
Returns
-------
PhysicalPropertyDataSet
"""
thermodynamic_state = ThermodynamicState(298.15 * unit.kelvin,
pressure=1.0 * unit.atmosphere)
source = MeasurementSource(doi="10.1000/xyz123")
solvation_substance = Substance()
solvation_substance.add_component(Component("O"), MoleFraction(1.0))
solvation_substance.add_component(Component("CCCO"), ExactAmount(1))
evaluator_properties = [
Density(
thermodynamic_state=thermodynamic_state,
phase=PropertyPhase.Liquid,
substance=Substance.from_components("O"),
value=1.0 * unit.kilogram / unit.meter**3,
uncertainty=0.1 * unit.kilogram / unit.meter**3,
source=source,
),
Density(
thermodynamic_state=thermodynamic_state,
phase=PropertyPhase.Liquid,
substance=Substance.from_components("O", "CC=O"),
value=1.0 * unit.kilogram / unit.meter**3,
uncertainty=0.1 * unit.kilogram / unit.meter**3,
source=source,
),
EnthalpyOfVaporization(
thermodynamic_state=thermodynamic_state,
phase=PropertyPhase(PropertyPhase.Liquid | PropertyPhase.Gas),
substance=Substance.from_components("CCO"),
value=1.0 * EnthalpyOfVaporization.default_unit(),
uncertainty=0.1 * EnthalpyOfVaporization.default_unit(),
source=source,
),
EnthalpyOfMixing(
thermodynamic_state=thermodynamic_state,
phase=PropertyPhase.Liquid,
substance=Substance.from_components("CCCCO", "CC(C=O)C"),
value=1.0 * EnthalpyOfMixing.default_unit(),
uncertainty=0.1 * EnthalpyOfMixing.default_unit(),
source=source,
),
ExcessMolarVolume(
thermodynamic_state=thermodynamic_state,
phase=PropertyPhase.Liquid,
substance=Substance.from_components("C(=O)CCCO", "CCCCCC"),
value=1.0 * ExcessMolarVolume.default_unit(),
uncertainty=0.1 * ExcessMolarVolume.default_unit(),
source=source,
),
SolvationFreeEnergy(
thermodynamic_state=thermodynamic_state,
phase=PropertyPhase.Liquid,
substance=solvation_substance,
value=1.0 * SolvationFreeEnergy.default_unit(),
uncertainty=0.1 * SolvationFreeEnergy.default_unit(),
source=source,
),
]
for index, evaluator_property in enumerate(evaluator_properties):
evaluator_property.id = str(index + 1)
evaluator_data_set = PhysicalPropertyDataSet()
evaluator_data_set.add_properties(*evaluator_properties)
return evaluator_data_set
def define_data_set(reweighting: bool) -> PhysicalPropertyDataSet:
# Define a common state to compute estimates at
states = [
ThermodynamicState(temperature=296.15 * unit.kelvin,
pressure=1.0 * unit.atmosphere),
ThermodynamicState(temperature=298.15 * unit.kelvin,
pressure=1.0 * unit.atmosphere),
ThermodynamicState(temperature=300.15 * unit.kelvin,
pressure=1.0 * unit.atmosphere),
]
data_set = PhysicalPropertyDataSet()
# Solvation free energies.
if not reweighting:
ethanol_substance = Substance.from_components("CCO")
ethanol_substance.add_component(
Component("CC=O", Component.Role.Solute), ExactAmount(1))
ethanal_substance = Substance.from_components("CC=O")
ethanal_substance.add_component(
Component("CCO", Component.Role.Solute), ExactAmount(1))
data_set.add_properties(
SolvationFreeEnergy(
thermodynamic_state=states[1],
phase=PropertyPhase.Liquid,
substance=ethanol_substance,
value=0.0 * SolvationFreeEnergy.default_unit(),
),
SolvationFreeEnergy(
thermodynamic_state=states[1],
phase=PropertyPhase.Liquid,
substance=ethanal_substance,
value=0.0 * SolvationFreeEnergy.default_unit(),
),
*CurationWorkflow.apply(
PhysicalPropertyDataSet(),
CurationWorkflowSchema(component_schemas=[
ImportFreeSolvSchema(),
FilterBySubstancesSchema(substances_to_include=[("O",
"CO")]),
]),
),
)
for state in states:
# Excess properties.
data_set.add_properties(
ExcessMolarVolume(
thermodynamic_state=state,
phase=PropertyPhase.Liquid,
substance=Substance.from_components("CC=O", "CCO"),
value=0.0 * ExcessMolarVolume.default_unit(),
),
EnthalpyOfMixing(
thermodynamic_state=state,
phase=PropertyPhase.Liquid,
substance=Substance.from_components("CC=O", "CCO"),
value=0.0 * EnthalpyOfMixing.default_unit(),
),
)
# Pure properties
data_set.add_properties(
Density(
thermodynamic_state=state,
phase=PropertyPhase.Liquid,
substance=Substance.from_components("CCO"),
value=0.0 * Density.default_unit(),
),
EnthalpyOfVaporization(
thermodynamic_state=state,
phase=PropertyPhase(PropertyPhase.Liquid | PropertyPhase.Gas),
substance=Substance.from_components("CCO"),
value=0.0 * EnthalpyOfVaporization.default_unit(),
),
DielectricConstant(
thermodynamic_state=state,
phase=PropertyPhase.Liquid,
substance=Substance.from_components("CCO"),
value=0.0 * DielectricConstant.default_unit(),
),
)
return data_set
def from_pandas(cls,
data_frame: pandas.DataFrame) -> "PhysicalPropertyDataSet":
"""Constructs a data set object from a pandas ``DataFrame`` object.
Notes
-----
* All physical properties are assumed to be source from experimental
measurements.
* Currently this method onlu supports data frames containing properties
which are built-in to the framework (e.g. Density).
* This method assumes the data frame has a structure identical to that
produced by the ``PhysicalPropertyDataSet.to_pandas`` function.
Parameters
----------
data_frame
The data frame to construct the data set from.
Returns
-------
The constructed data set.
"""
from openff.evaluator import properties
property_header_matches = {
re.match(r"^([a-zA-Z]+) Value \(([a-zA-Z0-9+-/\s]*)\)$", header)
for header in data_frame if header.find(" Value ") >= 0
}
property_headers = {}
# Validate that the headers have the correct format, specify a
# built-in property type, and specify correctly the properties
# units.
for match in property_header_matches:
assert match
property_type_string, property_unit_string = match.groups()
assert hasattr(properties, property_type_string)
property_type = getattr(properties, property_type_string)
property_unit = unit.Unit(property_unit_string)
assert property_unit is not None
assert (property_unit.dimensionality ==
property_type.default_unit().dimensionality)
property_headers[match.group(0)] = (property_type, property_unit)
# Convert the data rows to property objects.
physical_properties = []
for _, data_row in data_frame.iterrows():
data_row = data_row.dropna()
# Extract the state at which the measurement was made.
thermodynamic_state = ThermodynamicState(
temperature=data_row["Temperature (K)"] * unit.kelvin,
pressure=data_row["Pressure (kPa)"] * unit.kilopascal,
)
property_phase = PropertyPhase.from_string(data_row["Phase"])
# Extract the substance the measurement was made for.
substance = Substance()
for i in range(data_row["N Components"]):
component = Component(
smiles=data_row[f"Component {i + 1}"],
role=Component.Role[data_row.get(f"Role {i + 1}",
"Solvent")],
)
mole_fraction = data_row.get(f"Mole Fraction {i + 1}", 0.0)
exact_amount = data_row.get(f"Exact Amount {i + 1}", 0)
if not numpy.isclose(mole_fraction, 0.0):
substance.add_component(component,
MoleFraction(mole_fraction))
if not numpy.isclose(exact_amount, 0.0):
substance.add_component(component,
ExactAmount(exact_amount))
for (
property_header,
(property_type, property_unit),
) in property_headers.items():
# Check to see whether the row contains a value for this
# type of property.
if property_header not in data_row:
continue
uncertainty_header = property_header.replace(
"Value", "Uncertainty")
source_string = data_row["Source"]
is_doi = all(
any(
re.match(pattern, split_string, re.I) for pattern in [
r"^10.\d{4,9}/[-._;()/:A-Z0-9]+$",
r"^10.1002/[^\s]+$",
r"^10.\d{4}/\d+-\d+X?(\d+)\d+<[\d\w]+:[\d\w]*>\d+.\d+.\w+;\d$",
r"^10.1021/\w\w\d+$",
r"^10.1207/[\w\d]+\&\d+_\d+$",
]) for split_string in source_string.split(" + "))
physical_property = property_type(
thermodynamic_state=thermodynamic_state,
phase=property_phase,
value=data_row[property_header] * property_unit,
uncertainty=None if uncertainty_header not in data_row else
data_row[uncertainty_header] * property_unit,
substance=substance,
source=MeasurementSource(
doi="" if not is_doi else source_string,
reference=source_string if not is_doi else "",
),
)
identifier = data_row.get("Id", None)
if identifier:
physical_property.id = identifier
physical_properties.append(physical_property)
data_set = PhysicalPropertyDataSet()
data_set.add_properties(*physical_properties)
return data_set
def test_component_standardization(smiles, expected):
component = Component(smiles=smiles)
assert component.smiles == expected