def test_fragmentation_settings():
"""
Make sure the settings are correctly handled.
"""
fragmenter = workflow_components.WBOFragmenter()
with pytest.raises(ValueError):
fragmenter.functional_groups = get_data("functional_groups_error.yaml")
fragmenter.functional_groups = get_data("functional_groups.yaml")
assert fragmenter.functional_groups is not None
def test_wbo_fragmentation_apply():
"""
Make sure that wbo fragmentation is working.
"""
fragmenter = workflow_components.WBOFragmenter()
assert fragmenter.is_available()
# check that a molecule with no rotatable bonds fails if we dont want the parent back
benzene = Molecule.from_file(get_data("benzene.sdf"), "sdf")
result = fragmenter.apply([
benzene,
],
processors=1,
toolkit_registry=GLOBAL_TOOLKIT_REGISTRY)
assert result.n_molecules == 0
# now try a molecule which should give fragments
diphenhydramine = Molecule.from_smiles("O(CCN(C)C)C(c1ccccc1)c2ccccc2")
result = fragmenter.apply([
diphenhydramine,
],
processors=1,
toolkit_registry=GLOBAL_TOOLKIT_REGISTRY)
assert result.n_molecules == 4
for molecule in result.molecules:
assert "dihedrals" in molecule.properties
def test_create_torsiondrive_dataset():
"""
Make sure we can correclt make a dataset using the scan enumerator.
"""
factory = TorsiondriveDatasetFactory()
scan_filter = workflow_components.ScanEnumerator()
scan_filter.add_torsion_scan(smarts="[*:1]~[*:2]-[#8:3]-[#1:4]",
scan_rage=(-90, 90),
scan_increment=10)
factory.add_workflow_components(scan_filter)
conformer_generator = workflow_components.StandardConformerGenerator(
max_conformers=1)
factory.add_workflow_components(conformer_generator)
mols = Molecule.from_file(get_data("tautomers_small.smi"),
"smi",
allow_undefined_stereo=True)
dataset = factory.create_dataset(dataset_name="test name",
molecules=mols,
description="Force field test",
tagline="A test dataset",
processors=1)
assert dataset.n_molecules > 0
assert dataset.n_records > 0
for entry in dataset.dataset.values():
assert entry.keywords.dihedral_ranges == [(-90, 90)]
assert entry.keywords.grid_spacing == [10]
def test_create_dataset(factory_dataset_type):
"""
Test making a the correct corresponding dataset type from a given factory type.
"""
factory = factory_dataset_type[0]()
element_filter = workflow_components.ElementFilter()
element_filter.allowed_elements = [1, 6, 8, 7]
factory.add_workflow_component(element_filter)
conformer_generator = workflow_components.StandardConformerGenerator(max_conformers=1)
factory.add_workflow_component(conformer_generator)
mols = Molecule.from_file(get_data("tautomers_small.smi"), "smi", allow_undefined_stereo=True)
# set some settings
changed_attrs = {"compute_tag": "test tag",
"dataset_tags": ["openff", "test"],
"maxiter": 400}
for attr, value in changed_attrs.items():
setattr(factory, attr, value)
dataset = factory.create_dataset(dataset_name="test name", molecules=mols, description="Force field test", tagline="A test dataset")
# check the attributes were changed
for attr, value in changed_attrs.items():
assert getattr(dataset, attr) == value
assert dataset.dataset_name == "test name"
assert isinstance(dataset, factory_dataset_type[1]) is True
# make sure molecules we filtered and passed
assert dataset.dataset != {}
assert dataset.filtered != {}
assert element_filter.component_name in dataset.filtered_molecules
def get_stereoisomers():
"""
Get a set of molecules that all have some undefined stereochemistry.
"""
mols = Molecule.from_file(get_data("stereoisomers.smi"), allow_undefined_stereo=True)
return mols
def get_tautomers():
"""
Get a set of molecules that all have tautomers
"""
mols = Molecule.from_file(get_data("tautomers_small.smi"), allow_undefined_stereo=True)
return mols
def test_torsiondrive_linear_torsion():
"""
Test the torsiondrive factorys ability to find linear bonds which should not be driven.
"""
factory = TorsiondriveDatasetFactory()
molecules = Molecule.from_file(get_data("linear_molecules.sdf"), "sdf", allow_undefined_stereo=True)
for molecule in molecules:
assert bool(factory._detect_linear_torsions(molecule)) is True
def test_optimization_submissions_with_constraints(fractal_compute_server):
"""
Make sure that the constraints are added to the optimization and enforced.
"""
client = FractalClient(fractal_compute_server)
ethane = Molecule.from_file(get_data("ethane.sdf"), "sdf")
factory = OptimizationDatasetFactory()
dataset = OptimizationDataset(
dataset_name="Test optimizations with constraint",
description="Test optimization dataset with constraints",
tagline="Testing optimization datasets")
# add just mm spec
dataset.add_qc_spec(method="openff-1.0.0",
basis="smirnoff",
program="openmm",
spec_name="default",
spec_description="mm default spec",
overwrite=True)
# build some constraints
constraints = Constraints()
constraints.add_set_constraint(constraint_type="dihedral",
indices=[2, 0, 1, 5],
value=60,
bonded=True)
constraints.add_freeze_constraint(constraint_type="distance",
indices=[0, 1],
bonded=True)
# add the molecule
attributes = factory.create_cmiles_metadata(ethane)
index = ethane.to_smiles()
dataset.add_molecule(index=index,
molecule=ethane,
attributes=attributes,
constraints=constraints)
# now add a mock url so we can submit the data
dataset.metadata.long_description_url = "https://test.org"
# now submit again
dataset.submit(client=client)
fractal_compute_server.await_results()
# make sure of the results are complete
ds = client.get_collection("OptimizationDataset", dataset.dataset_name)
record = ds.get_record(ds.df.index[0], "default")
assert "constraints" in record.keywords
assert record.status.value == "COMPLETE"
assert record.error is None
assert len(record.trajectory) > 1
# now make sure the constraints worked
final_molecule = record.get_final_molecule()
assert pytest.approx(60, final_molecule.measure((2, 0, 1, 5)))
assert pytest.approx(record.get_initial_molecule().measure((0, 1)),
final_molecule.measure((0, 1)))
def test_serializer_round_trips(serializer):
"""
Test serializing data to and from file with no compression.
"""
# get data in a dict format
data = deserialize(get_data("settings_with_workflow.json"))
file_name = "settings_with_workflow" + serializer
# now export to file and back
with temp_directory():
serialize(serializable=data, file_name=file_name, compression=None)
deserialized_data = deserialize(file_name=file_name)
assert data == deserialized_data
def test_fragmentation_apply():
"""
Make sure that fragmentation is working.
"""
fragmenter = workflow_components.WBOFragmenter()
assert fragmenter.is_available()
# check that a molecule with no rotatable bonds fails if we dont want the parent back
benzene = Molecule.from_file(get_data("benzene.sdf"), "sdf")
result = fragmenter.apply([benzene, ], processors=1)
assert result.n_molecules == 0
# now try ethanol
ethanol = Molecule.from_file(get_data("methanol.sdf"), "sdf")
fragmenter.include_parent = True
result = fragmenter.apply([ethanol, ], processors=1)
assert result.n_molecules == 1
# now try a molecule which should give fragments
diphenhydramine = Molecule.from_smiles("O(CCN(C)C)C(c1ccccc1)c2ccccc2")
fragmenter.include_parent = False
result = fragmenter.apply([diphenhydramine, ], processors=1)
assert result.n_molecules == 4
for molecule in result.molecules:
assert "dihedrals" in molecule.properties
def test_compression_serialization_round_trip_file_name(
serialization, compression):
"""
Test all of the different serialization and compression combinations.
Here the compression is in the file name.
"""
# get data in a dict format
data = deserialize(get_data("settings_with_workflow.json"))
file_name = "".join(
["settings_with_workflow", ".", serialization, ".", compression])
# now export the file and read back
with temp_directory():
serialize(serializable=data, file_name=file_name, compression=None)
deserialized_data = deserialize(file_name=file_name)
assert data == deserialized_data
def test_torsiondrive_index():
"""
Test generating an index using torsiondrive, this should tag the atoms in the torsion.
"""
mol = Molecule.from_file(get_data("methanol.sdf"))
mol.properties["atom_map"] = {4: 0, 0: 1, 1: 2, 5: 3}
factory = TorsiondriveDatasetFactory()
index = factory.create_index(mol)
tags = ["[C:2]", "[H:1]", "[O:3]", "[H:4]"]
for tag in tags:
assert tag in index
def test_importing_settings_no_workflow(file_type, factory_type):
"""
Test importing the settings with no workflow components from the supported file types.
"""
factory = factory_type()
file_name = "settings." + file_type
factory.import_settings(get_data(file_name))
changed_attrs = {
"priority": "super_high",
"compute_tag": "loaded tag",
}
for attr, value in changed_attrs.items():
assert getattr(factory, attr) == value
def test_import_workflow_only(file_type, factory_type):
"""
Test importing a workflow only from a workflow file.
"""
factory = factory_type()
factory2 = factory_type()
file_name = "settings_with_workflow." + file_type
factory.import_workflow(get_data(file_name))
# make sure the settings have not changed from default
assert factory.dict(exclude={"workflow"}) == factory2.dict(exclude={"workflow"})
assert len(factory.workflow) == 1
assert factory.workflow != factory2.workflow
def test_improper_enumerator():
"""
Make sure improper torsions are correctly tagged.
"""
mol = Molecule.from_file(get_data("benzene.sdf"))
scan_tagger = workflow_components.ScanEnumerator()
# even though there is more than one improper make sure we only get one scan back
scan_tagger.add_improper_torsion(smarts="[#6:1](-[#1:2])(:[#6:3]):[#6:4]",
central_smarts="[#6:1]",
scan_range=(-40, 40),
scan_increment=4)
result = scan_tagger.apply([mol],
processors=1,
toolkit_registry=GLOBAL_TOOLKIT_REGISTRY)
assert result.n_molecules == 1
indexer = mol.properties["dihedrals"]
assert indexer.n_impropers == 1
assert indexer.impropers[0].scan_increment == [4]
def test_scan_filter_exclude():
"""
Make sure only exclude torsion scans we do not want and keep any other defined scans.
"""
ethanol = Molecule.from_file(get_data("ethanol.sdf"))
t_indexer = TorsionIndexer()
# C-C torsion
t_indexer.add_torsion(torsion=(3, 0, 1, 2), symmetry_group=(1, 1))
# O-C torsion
t_indexer.add_torsion(torsion=(0, 1, 2, 8), symmetry_group=(1, 2))
ethanol.properties["dihedrals"] = t_indexer
# only keep the C-C scan
filter = workflow_components.ScanFilter(
scans_to_exclude=["[#1:1]~[#6:2]-[#6:3]-[#1:4]"])
result = filter.apply(molecules=[ethanol],
processors=1,
toolkit_registry=GLOBAL_TOOLKIT_REGISTRY)
assert result.n_molecules == 1
assert (1, 2) in ethanol.properties["dihedrals"].torsions
def test_torsiondrive_torsion_string():
"""
Test the torsiondrive factories ability to create a torsion string for a given bond.
"""
methanol = Molecule.from_file(get_data("methanol.sdf"), "sdf")
rotatable = methanol.find_rotatable_bonds()
assert len(rotatable) == 1
bond = rotatable[0]
torsion = get_torsion(bond=bond)
# now make sure this torsion is in the propers list
reference_torsions = []
for proper in methanol.propers:
dihedral = []
for atom in proper:
dihedral.append(atom.molecule_atom_index)
reference_torsions.append(tuple(dihedral))
assert torsion in reference_torsions or tuple(
reversed(torsion)) in reference_torsions
def test_importing_settings_workflow(file_type, factory_type):
"""
Test importing the settings and a workflow from the supported file types.
"""
factory = factory_type()
file_name = "settings_with_workflow." + file_type
factory.import_settings(get_data(file_name))
changed_attrs = {
"maxiter": 400,
"priority": "super_high",
"compute_tag": "loaded tag",
}
for attr, value in changed_attrs.items():
assert getattr(factory, attr) == value
assert len(factory.workflow) == 1
assert "StandardConformerGenerator" in factory.workflow
component = factory.get_workflow_component("StandardConformerGenerator")
assert component.component_description == "loaded component"
assert isinstance(component, workflow_components.StandardConformerGenerator) is True
def test_basic_submissions_multiple_spec(fractal_compute_server):
"""Test submitting a basic dataset to a snowflake server with multiple qcspecs."""
client = FractalClient(fractal_compute_server)
qc_specs = [{
"method": "openff-1.0.0",
"basis": "smirnoff",
"program": "openmm",
"spec_name": "openff"
}, {
"method": "gaff-2.11",
"basis": "antechamber",
"program": "openmm",
"spec_name": "gaff"
}]
molecules = Molecule.from_file(get_data("butane_conformers.pdb"), "pdb")
factory = BasicDatasetFactory(driver="energy")
factory.clear_qcspecs()
for spec in qc_specs:
factory.add_qc_spec(**spec,
spec_description="testing the single points")
dataset = factory.create_dataset(
dataset_name="Test single points multiple specs",
molecules=molecules,
description="Test basics dataset",
tagline="Testing single point datasets",
)
# force a metadata validation error
dataset.metadata.long_description = None
with pytest.raises(DatasetInputError):
dataset.submit(client=client)
# re-add the description so we can submit the data
dataset.metadata.long_description = "Test basics dataset"
# now submit again
dataset.submit(client=client)
fractal_compute_server.await_results()
# make sure of the results are complete
ds = client.get_collection("Dataset", dataset.dataset_name)
# check the metadata
meta = Metadata(**ds.data.metadata)
assert meta == dataset.metadata
assert ds.data.description == dataset.description
assert ds.data.tagline == dataset.dataset_tagline
assert ds.data.tags == dataset.dataset_tags
# check the provenance
assert dataset.provenance == ds.data.provenance
# check the qc spec
assert ds.data.default_driver == dataset.driver
# get the last ran spec
for specification in ds.data.history:
driver, program, method, basis, spec_name = specification
spec = dataset.qc_specifications[spec_name]
assert driver == dataset.driver
assert program == spec.program
assert method == spec.method
assert basis == spec.basis
for spec in dataset.qc_specifications.values():
query = ds.get_records(
method=spec.method,
basis=spec.basis,
program=spec.program,
)
# make sure all conformers are submitted
assert len(query.index) == len(molecules)
for index in query.index:
result = query.loc[index].record
assert result.status.value.upper() == "COMPLETE"
assert result.error is None
assert result.return_result is not None
def test_optimization_submissions(fractal_compute_server, specification):
"""Test submitting an Optimization dataset to a snowflake server."""
client = FractalClient(fractal_compute_server)
qc_spec, driver = specification
program = qc_spec["program"]
if not has_program(program):
pytest.skip(f"Program '{program}' not found.")
molecules = Molecule.from_file(get_data("butane_conformers.pdb"), "pdb")
factory = OptimizationDatasetFactory(driver=driver)
factory.add_qc_spec(**qc_spec,
spec_name="default",
spec_description="test",
overwrite=True)
dataset = factory.create_dataset(
dataset_name=f"Test optimizations info {program}, {driver}",
molecules=molecules[:2],
description="Test optimization dataset",
tagline="Testing optimization datasets",
)
# force a metadata validation error
dataset.metadata.long_description = None
with pytest.raises(DatasetInputError):
dataset.submit(client=client)
# re-add the description so we can submit the data
dataset.metadata.long_description = "Test basics dataset"
# now submit again
dataset.submit(client=client)
fractal_compute_server.await_results()
# make sure of the results are complete
ds = client.get_collection("OptimizationDataset", dataset.dataset_name)
# check the metadata
meta = Metadata(**ds.data.metadata)
assert meta == dataset.metadata
# check the provenance
assert dataset.provenance == ds.data.provenance
# check the qc spec
for qc_spec in dataset.qc_specifications.values():
spec = ds.data.specs[qc_spec.spec_name]
assert spec.description == qc_spec.spec_description
assert spec.qc_spec.driver == dataset.driver
assert spec.qc_spec.method == qc_spec.method
assert spec.qc_spec.basis == qc_spec.basis
assert spec.qc_spec.program == qc_spec.program
# check the keywords
keywords = client.query_keywords(spec.qc_spec.keywords)[0]
assert keywords.values["maxiter"] == qc_spec.maxiter
assert keywords.values["scf_properties"] == qc_spec.scf_properties
# query the dataset
ds.query(qc_spec.spec_name)
for index in ds.df.index:
record = ds.df.loc[index].default
assert record.status.value == "COMPLETE"
assert record.error is None
assert len(record.trajectory) > 1
# if we used psi4 make sure the properties were captured
if program == "psi4":
result = record.get_trajectory()[0]
assert "CURRENT DIPOLE X" in result.extras["qcvars"].keys()
assert "SCF QUADRUPOLE XX" in result.extras["qcvars"].keys()
def test_basic_submissions_single_spec(fractal_compute_server, specification):
"""Test submitting a basic dataset to a snowflake server."""
client = FractalClient(fractal_compute_server)
qc_spec, driver = specification
program = qc_spec["program"]
if not has_program(program):
pytest.skip(f"Program '{program}' not found.")
molecules = Molecule.from_file(get_data("butane_conformers.pdb"), "pdb")
factory = BasicDatasetFactory(driver=driver)
factory.add_qc_spec(**qc_spec,
spec_name="default",
spec_description="testing the single points",
overwrite=True)
dataset = factory.create_dataset(
dataset_name=f"Test single points info {program}, {driver}",
molecules=molecules,
description="Test basics dataset",
tagline="Testing single point datasets",
)
# force a metadata validation error
dataset.metadata.long_description = None
with pytest.raises(DatasetInputError):
dataset.submit(client=client)
# re-add the description so we can submit the data
dataset.metadata.long_description = "Test basics dataset"
# now submit again
dataset.submit(client=client)
fractal_compute_server.await_results()
# make sure of the results are complete
ds = client.get_collection("Dataset", dataset.dataset_name)
# check the metadata
meta = Metadata(**ds.data.metadata)
assert meta == dataset.metadata
assert ds.data.description == dataset.description
assert ds.data.tagline == dataset.dataset_tagline
assert ds.data.tags == dataset.dataset_tags
# check the provenance
assert dataset.provenance == ds.data.provenance
# check the qc spec
assert ds.data.default_driver == dataset.driver
# get the last ran spec
for specification in ds.data.history:
driver, program, method, basis, spec_name = specification
spec = dataset.qc_specifications[spec_name]
assert driver == dataset.driver
assert program == spec.program
assert method == spec.method
assert basis == spec.basis
break
else:
raise RuntimeError(
f"The requested compute was not found in the history {ds.data.history}"
)
for spec in dataset.qc_specifications.values():
query = ds.get_records(
method=spec.method,
basis=spec.basis,
program=spec.program,
)
# make sure all of the conformers were submitted
assert len(query.index) == len(molecules)
for index in query.index:
result = query.loc[index].record
assert result.status.value.upper() == "COMPLETE"
assert result.error is None
assert result.return_result is not None
def test_basic_submissions_wavefunction(fractal_compute_server):
"""
Test submitting a basic dataset with a wavefunction protocol and make sure it is executed.
"""
# only a psi4 test
if not has_program("psi4"):
pytest.skip("Program psi4 not found.")
client = FractalClient(fractal_compute_server)
molecules = Molecule.from_file(get_data("butane_conformers.pdb"), "pdb")
factory = BasicDatasetFactory(driver="energy")
factory.clear_qcspecs()
factory.add_qc_spec(method="hf",
basis="sto-6g",
program="psi4",
spec_name="default",
spec_description="wavefunction spec",
store_wavefunction="orbitals_and_eigenvalues")
dataset = factory.create_dataset(
dataset_name="Test single points with wavefunction",
molecules=molecules,
description="Test basics dataset",
tagline="Testing single point datasets with wavefunction",
)
# submit the dataset
# now submit again
dataset.submit(client=client)
fractal_compute_server.await_results()
# make sure of the results are complete
ds = client.get_collection("Dataset", dataset.dataset_name)
# check the metadata
meta = Metadata(**ds.data.metadata)
assert meta == dataset.metadata
assert ds.data.description == dataset.description
assert ds.data.tagline == dataset.dataset_tagline
assert ds.data.tags == dataset.dataset_tags
# check the provenance
assert dataset.provenance == ds.data.provenance
# check the qc spec
assert ds.data.default_driver == dataset.driver
# get the last ran spec
for specification in ds.data.history:
driver, program, method, basis, spec_name = specification
spec = dataset.qc_specifications[spec_name]
assert driver == dataset.driver
assert program == spec.program
assert method == spec.method
assert basis == spec.basis
for spec in dataset.qc_specifications.values():
query = ds.get_records(
method=spec.method,
basis=spec.basis,
program=spec.program,
)
for index in query.index:
result = query.loc[index].record
assert result.status.value.upper() == "COMPLETE"
assert result.error is None
assert result.return_result is not None
basis = result.get_wavefunction("basis")
assert basis.name.lower() == "sto-6g"
orbitals = result.get_wavefunction("orbitals_a")
assert orbitals.shape is not None
def test_basic_submissions_single_pcm_spec(fractal_compute_server):
"""Test submitting a basic dataset to a snowflake server with pcm water in the specification."""
client = FractalClient(fractal_compute_server)
program = "psi4"
if not has_program(program):
pytest.skip(f"Program '{program}' not found.")
molecules = Molecule.from_file(get_data("butane_conformers.pdb"), "pdb")
factory = BasicDatasetFactory(driver="energy")
factory.add_qc_spec(method="hf",
basis="sto-3g",
program=program,
spec_name="default",
spec_description="testing the single points with pcm",
implicit_solvent=PCMSettings(units="au",
medium_Solvent="water"),
overwrite=True)
# only use one molecule due to the time it takes to run with pcm
dataset = factory.create_dataset(
dataset_name="Test single points with pcm water",
molecules=molecules[0],
description="Test basics dataset with pcm water",
tagline="Testing single point datasets with pcm water",
)
# force a metadata validation error
dataset.metadata.long_description = None
with pytest.raises(DatasetInputError):
dataset.submit(client=client)
# re-add the description so we can submit the data
dataset.metadata.long_description = "Test basics dataset"
# now submit again
dataset.submit(client=client)
fractal_compute_server.await_results()
# make sure of the results are complete
ds = client.get_collection("Dataset", dataset.dataset_name)
# check the metadata
meta = Metadata(**ds.data.metadata)
assert meta == dataset.metadata
assert ds.data.description == dataset.description
assert ds.data.tagline == dataset.dataset_tagline
assert ds.data.tags == dataset.dataset_tags
# check the provenance
assert dataset.provenance == ds.data.provenance
# check the qc spec
assert ds.data.default_driver == dataset.driver
# get the last ran spec
for specification in ds.data.history:
driver, program, method, basis, spec_name = specification
spec = dataset.qc_specifications[spec_name]
assert driver == dataset.driver
assert program == spec.program
assert method == spec.method
assert basis == spec.basis
break
else:
raise RuntimeError(
f"The requested compute was not found in the history {ds.data.history}"
)
for spec in dataset.qc_specifications.values():
query = ds.get_records(
method=spec.method,
basis=spec.basis,
program=spec.program,
)
for index in query.index:
result = query.loc[index].record
assert result.status.value.upper() == "COMPLETE"
assert result.error is None
assert result.return_result is not None
# make sure the PCM result was captured
assert result.extras["qcvars"]["PCM POLARIZATION ENERGY"] < 0
