def create_csv(path: Union[str, PathLike], database: Union[Ligand, QD] = 'ligand') -> str:
"""Create a ligand or qd database (csv format) if it does not yet exist.
Parameters
----------
path : str
The path (without filename) of the database.
database : |str|_
The type of database, accepted values are ``"ligand"`` and ``"qd"``.
Returns
-------
|str|_
The absolute path to the ligand or qd database.
"""
filename = join(path, f'{database}_database.csv')
# Check if the database exists and has the proper keys; create it if it does not
if not isfile(filename):
if database == 'ligand':
_create_csv_lig(filename)
elif database == 'qd':
_create_csv_qd(filename)
else:
raise ValueError(f"{database!r} is not an accepated value for the 'database' argument")
logger.info(f'{database}_database.csv not found in {path}, creating {database} database')
return filename
def _crs_run(job: CRSJob,
name: str,
calc_type: str = 'activity coefficient') -> CRSResults:
"""Call the :meth:`CRSJob.run` on **job**, returning a :class:`CRSResults` instance."""
_name = _get_name(job.name)
logger.info(f'{job.__class__.__name__}: {name} {calc_type} calculation '
f'({_name}) has started')
return job.run(jobrunner=JobRunner(parallel=True))
def _update_index_dset(group: h5py.Group, name: str, logger: Optional[Logger] = None) -> None:
"""Check for and update pre dataCAT 0.4 style databases."""
if 'index' in group:
return None
elif logger is not None:
logger.info(f'Updating h5py Dataset to data-CAT >= 0.4 style: {name!r}')
dtype = IDX_DTYPE[name]
i = len(group['atoms'])
_set_index(PDBContainer, group, dtype, i, compression='gzip')
def get_pka(mol: Molecule, coskf_mol: Optional[str],
coskf_mol_conj: Optional[str], water: str, hydronium: str,
job: Type[Job], s: Settings) -> List[float]:
if coskf_mol is None:
return 5 * [np.nan]
elif coskf_mol_conj is None:
return 5 * [np.nan]
s = Settings(s)
s.input.compound[1]._h = water
s.ignore_molecule = True
s_dict = {}
for name, coskf in _iter_coskf(coskf_mol_conj, coskf_mol, water,
hydronium):
_s = s.copy()
_s.name = name
_s.input.compound[0]._h = coskf
s_dict[name] = _s
# Run the job
mol_name = mol.properties.name
job_list = [CRSJob(settings=s, name=name) for name, s in s_dict.items()]
results_list = [_crs_run(job, mol_name) for job in job_list]
# Extract solvation energies and activity coefficients
E_solv = {}
for name, results in zip(("acid", "base", "solvent", "solvent_conj"),
results_list):
results.wait()
try:
E_solv[name] = _E = results.get_energy()
assert _E is not None
logger.info(f'{results.job.__class__.__name__}: {mol_name} pKa '
f'calculation ({results.job.name}) is successful')
except Exception:
logger.error(f'{results.job.__class__.__name__}: {mol_name} pKa '
f'calculation ({results.job.name}) has failed')
E_solv[name] = np.nan
try:
mol.properties.job_path += [
join(job.path, job.name + '.in') for job in job_list
]
except IndexError: # The 'job_path' key is not available
mol.properties.job_path = [
join(job.path, job.name + '.in') for job in job_list
]
ret = [E_solv[k] for k in ("acid", "base", "solvent", "solvent_conj")]
ret.append(_get_pka(**E_solv))
return ret
def _get_logp(s: Settings, name: str, logger: logging.Logger) -> float:
logp_s = s.copy()
logp_s.update(get_template('qd.yaml')['COSMO-RS logp'])
for v in logp_s.input.compound:
v._h = v._h.format(os.environ["ADFRESOURCES"])
logp_job = CRSJob(settings=logp_s, name='LogP')
results = _crs_run(logp_job, name)
try:
logp = results.readkf('LOGP', 'logp')[0]
logger.info(f'{results.job.__class__.__name__}: {name} LogP '
f'calculation ({results.job.name}) is successful')
except Exception:
logger.error(f'{results.job.__class__.__name__}: {name} LogP '
f'calculation ({results.job.name}) has failed')
logp = np.nan
return logp
def _update_pdb_dsets(file: h5py.File, name: str,
logger: Optional[Logger] = None) -> Optional[PDBContainer]:
"""Check for and update pre dataCAT 0.3 style databases."""
if not isinstance(file.get(name), h5py.Dataset):
return None
elif logger is not None:
logger.info(f'Updating h5py Dataset to data-CAT >= 0.3 style: {name!r}')
mol_list = [from_pdb_array(pdb, rdmol=False, warn=False) for pdb in file[name]]
m = len(mol_list)
del file[name]
dtype = IDX_DTYPE[name]
scale = np.rec.array(None, dtype=dtype, shape=(m,))
if dtype.fields is not None and scale.size:
# Ensure that the sentinal value for vlen strings is an empty string, not `None`
elem = list(scale.item(0))
iterator = (v for v, *_ in dtype.fields.values())
for i, sub_dt in enumerate(iterator):
if h5py.check_string_dtype(sub_dt) is not None:
elem[i] = ''
scale[:] = tuple(elem)
return PDBContainer.from_molecules(mol_list, scale=scale)
def get_solv(mol: Molecule, solvent_list: Iterable[str], coskf: Optional[str],
job: Type[Job], s: Settings) -> List[float]:
"""Calculate the solvation energy of *mol* in various *solvents*.
Parameters
----------
mol : |plams.Molecule|_
A PLAMS Molecule.
solvent_list : |List|_ [|str|_]
A list of solvent molecules (*i.e.* .coskf files).
coskf : str, optional
The path+filename of the .coskf file of **mol**.
job : |Callable|_
A type Callable of a class derived from :class:`Job`, e.g. :class:`AMSJob`
or :class:`Cp2kJob`.
s : |plams.Settings|_
The settings for **job**.
Returns
-------
|list|_ [|float|_] & |list|_ [|float|_]
A list of solvation energies and gammas.
"""
# Return 3x np.nan if no coskf is None (i.e. the COSMO-surface construction failed)
if coskf is None:
i = 1 + 2 * len(solvent_list)
return i * [np.nan]
# Prepare a list of job settings
s = Settings(s)
s.input.compound[0]._h = coskf
s.ignore_molecule = True
s_list = []
for solv in solvent_list:
_s = s.copy()
_s.name = solv.rsplit('.', 1)[0].rsplit(os.sep, 1)[-1]
_s.input.compound[1]._h = solv
s_list.append(_s)
# Run the job
mol_name = mol.properties.name
job_list = [CRSJob(settings=s, name=s.name) for s in s_list]
results_list = [
_crs_run(job, mol_name, calc_type="pKa") for job in job_list
]
# Extract solvation energies and activity coefficients
E_solv = []
Gamma = []
for results in results_list:
results.wait()
try:
E_solv.append(results.get_energy())
Gamma.append(results.get_activity_coefficient())
logger.info(
f'{results.job.__class__.__name__}: {mol_name} activity coefficient '
f'calculation ({results.job.name}) is successful')
except Exception:
logger.error(
f'{results.job.__class__.__name__}: {mol_name} activity coefficient '
f'calculation ({results.job.name}) has failed')
E_solv.append(np.nan)
Gamma.append(np.nan)
try:
mol.properties.job_path += [
join(job.path, job.name + '.in') for job in job_list
]
except IndexError: # The 'job_path' key is not available
mol.properties.job_path = [
join(job.path, job.name + '.in') for job in job_list
]
logp = _get_logp(s_list[0], name=mol_name, logger=logger)
return E_solv + Gamma + [logp]
def run_match_job(mol: Molecule,
s: Settings,
job_type: Type[Job] = MatchJob,
action: str = 'warn') -> None:
"""Assign atom types and charges to **mol** based on the results of MATCH_.
Performs an inplace update of :attr:`Atom.properties` ``["symbol"]``,
:attr:`Atom.properties` ``["charge"]`` and :attr:`Molecule.properties` ``["prm"]``.
.. _MATCH: http://brooks.chem.lsa.umich.edu/index.php?page=match&subdir=articles/resources/software
Parameters
----------
mol : |plams.Molecule|
A PLAMS molecule.
s : |plams.Settings|
Job settings for the to-be constructed :class:`.MatchJob` instance.
job_type : :class:`type` [|plams.Job|]
The type of Job.
action : :class:`str`
The to-be undertaken action when the Job crashes.
Accepted values are ``"raise"``, ``"warn"`` and ``"ignore"``.
See Also
--------
:class:`.MatchJob`
A :class:`Job` subclass for interfacing with MATCH_: Multipurpose Atom-Typer for CHARMM.
""" # noqa
job = job_type(molecule=mol, settings=s, name='ff_assignment')
# Run the job
try:
results = job.run()
if job.status != 'successful':
raise _get_results_error(results)
symbol_list = results.get_atom_types()
charge_list = results.get_atom_charges()
logger.info(
f'{job.__class__.__name__}: {mol.properties.name} parameter assignment '
f'({job.name}) is successful')
except Exception as ex:
if action == 'raise':
raise ex
elif 'action' == 'warm':
logger.info(
f'{job.__class__.__name__}: {mol.properties.name} parameter assignment '
f'({job.name}) has failed')
logger.debug(f'{ex.__class__.__name__}: {ex}', exc_info=True)
return None
# Update properties with new symbols, charges and the consntructed parameter (.prm) file
mol.properties.prm = prm = results['$JN.prm']
for at, symbol, charge in zip(mol, symbol_list, charge_list):
at.properties.symbol = symbol
at.properties.charge_float = charge
post_proccess_prm(prm)
return None