datamol.descriptors

n_aromatic_atoms(mol)

Calculate the number of aromatic atoms.

Source code in datamol/descriptors/descriptors.py

def n_aromatic_atoms(mol: Mol) -> int:
    """Calculate the number of aromatic atoms."""
    matches = mol.GetSubstructMatches(_AROMATIC_QUERY)
    return len(matches)

n_aromatic_atoms_proportion(mol)

Calculate the aromatic proportion: # aromatic atoms/#atoms total.

Parameters:

Name	Type	Description	Default
mol	Mol	A molecule.	required

Only heavy atoms are considered.

Source code in datamol/descriptors/descriptors.py

def n_aromatic_atoms_proportion(mol: Mol) -> int:
    """Calculate the aromatic proportion: # aromatic atoms/#atoms total.

    Args:
        mol: A molecule.

    Only heavy atoms are considered.
    """
    return n_aromatic_atoms(mol) / mol.GetNumHeavyAtoms()

n_charged_atoms(mol)

Compute the number of charged atoms in a molecule.

Parameters:

Name	Type	Description	Default
mol	Mol	A molecule.	required

Returns:

Name	Type	Description
n_charged_atoms	int	number of charged atoms in the molecule

Source code in datamol/descriptors/descriptors.py

def n_charged_atoms(mol: Mol) -> int:
    """Compute the number of charged atoms in a molecule.

    Args:
        mol: A molecule.

    Returns:
        n_charged_atoms: number of charged atoms in the molecule
    """
    return sum([at.GetFormalCharge() != 0 for at in mol.GetAtoms()])

n_rigid_bonds(mol)

Compute the number of rigid bonds in a molecule.

Rigid bonds are bonds that are not single and not in rings.

Parameters:

Name	Type	Description	Default
mol	Mol	A molecule.	required

Returns:

Name	Type	Description
n_rigid_bonds	int	number of rigid bonds in the molecule

Source code in datamol/descriptors/descriptors.py

def n_rigid_bonds(mol: Mol) -> int:
    """Compute the number of rigid bonds in a molecule.

    Rigid bonds are bonds that are not single and not in rings.

    Args:
        mol: A molecule.

    Returns:
        n_rigid_bonds: number of rigid bonds in the molecule
    """
    non_rigid_bonds_count = from_smarts("*-&!@*")
    n_rigid_bonds = mol.GetNumBonds() - len(mol.GetSubstructMatches(non_rigid_bonds_count))
    return n_rigid_bonds

n_stereo_centers(mol)

Compute the number of stereocenters in a molecule.

Parameters:

Name	Type	Description	Default
mol	Mol	A molecule.	required

Returns:

Name	Type	Description
n_stero_center	int	number of stereocenters in the molecule

Source code in datamol/descriptors/descriptors.py

def n_stereo_centers(mol: Mol) -> int:
    """Compute the number of stereocenters in a molecule.

    Args:
        mol: A molecule.

    Returns:
        n_stero_center: number of stereocenters in the molecule
    """
    n = 0
    try:
        rdmolops.FindPotentialStereo(mol, cleanIt=False)
        n = rdMolDescriptors.CalcNumAtomStereoCenters(mol)
    except Exception:
        pass
    return n

n_stereo_centers_unspecified(mol)

Compute the number of unspecified stereocenters in a molecule.

Parameters:

Name	Type	Description	Default
mol	Mol	A molecule.	required

Returns:

Name	Type	Description
n_stereo_centers_unspecified	int	number of unspecified stereocenters in the molecule

Source code in datamol/descriptors/descriptors.py

def n_stereo_centers_unspecified(mol: Mol) -> int:
    """Compute the number of unspecified stereocenters in a molecule.

    Args:
        mol: A molecule.

    Returns:
        n_stereo_centers_unspecified: number of unspecified stereocenters in the molecule
    """
    n = 0
    try:
        rdmolops.FindPotentialStereo(mol, cleanIt=False)
        n = rdMolDescriptors.CalcNumUnspecifiedAtomStereoCenters(mol)
    except Exception:
        pass
    return n

any_rdkit_descriptor(name)

Return a descriptor function by name either from rdkit.Chem import Descriptors or rdkit.Chem.rdMolDescriptors.

Parameters:

Name	Type	Description	Default
name	str	Descriptor name.	required

Source code in datamol/descriptors/compute.py

def any_rdkit_descriptor(name: str) -> Callable:
    """Return a descriptor function by name either from
    `rdkit.Chem import Descriptors` or `rdkit.Chem.rdMolDescriptors`.

    Args:
        name: Descriptor name.
    """
    fn = getattr(Descriptors, name, None)

    if fn is None:
        fn = getattr(rdMolDescriptors, name, None)

    if fn is None:
        raise ValueError(f"Descriptor {name} not found.")

    return fn

batch_compute_many_descriptors(mols, properties_fn=None, add_properties=True, n_jobs=1, batch_size=None, progress=False, progress_leave=True)

Compute a list of opiniated molecular properties on a list of molecules.

Parameters:

Name	Type	Description	Default
mols	List[Mol]	A list of molecules.	required
properties_fn	Optional[Dict[str, Union[Callable, str]]]	A list of functions that compute properties. If None, a default list of properties is used. If the function is a string, dm.descriptors.any_descriptor() is used to retrieve the descriptor function.	None
add_properties	bool	Whether to add the computed properties to the default list.	True

Returns:

Type	Description
DataFrame	A dataframe of computed properties with one row per input molecules.

Source code in datamol/descriptors/compute.py

def batch_compute_many_descriptors(
    mols: List[Mol],
    properties_fn: Optional[Dict[str, Union[Callable, str]]] = None,
    add_properties: bool = True,
    n_jobs: int = 1,
    batch_size: Optional[int] = None,
    progress: bool = False,
    progress_leave: bool = True,
) -> pd.DataFrame:
    """Compute a list of opiniated molecular properties on a list of molecules.

    Args:
        mols: A list of molecules.
        properties_fn: A list of functions that compute properties. If None,
            a default list of properties is used. If the function is a string,
            `dm.descriptors.any_descriptor()` is used to retrieve the descriptor
            function.
        add_properties: Whether to add the computed properties to the default list.

    Returns:
        A dataframe of computed properties with one row per input molecules.
    """

    compute_fn = functools.partial(
        compute_many_descriptors,
        properties_fn=properties_fn,
        add_properties=add_properties,
    )

    props = parallelized(
        compute_fn,
        mols,
        batch_size=batch_size,
        progress=progress,
        n_jobs=n_jobs,
        tqdm_kwargs=dict(leave=progress_leave),
    )
    return pd.DataFrame(props)

compute_many_descriptors(mol, properties_fn=None, add_properties=True)

Compute a list of opiniated molecular properties.

Parameters:

Name	Type	Description	Default
mol	Mol	A molecule.	required
properties_fn	Optional[Dict[str, Union[Callable, str]]]	A list of functions that compute properties. If None, a default list of properties is used. If the function is a string, dm.descriptors.any_descriptor() is used to retrieve the descriptor function.	None
add_properties	bool	Whether to add the computed properties to the default list.	True

Returns:

Type	Description
dict	Computed properties as a dict.

Source code in datamol/descriptors/compute.py

def compute_many_descriptors(
    mol: Mol,
    properties_fn: Optional[Dict[str, Union[Callable, str]]] = None,
    add_properties: bool = True,
) -> dict:
    """Compute a list of opiniated molecular properties.

    Args:
        mol: A molecule.
        properties_fn: A list of functions that compute properties. If None,
            a default list of properties is used. If the function is a string,
            `dm.descriptors.any_descriptor()` is used to retrieve the descriptor
            function.
        add_properties: Whether to add the computed properties to the default list.

    Returns:
        Computed properties as a dict.
    """

    if properties_fn is None:
        properties_fn = _DEFAULT_PROPERTIES_FN
    elif add_properties:
        [properties_fn.setdefault(k, v) for k, v in _DEFAULT_PROPERTIES_FN.items()]

    props = {}
    for k, v in properties_fn.items():
        if isinstance(v, str):
            v = any_rdkit_descriptor(v)

        props[k] = v(mol)

    return props