rdtools.dist#

A module contains functions to manipulate distances in a molecule.

rdtools.dist.get_adjacency_matrix(mol: Mol) → ndarray[tuple[int, ...], dtype[int64]]#

Get the adjacency matrix of the molecule.

Parameters:: mol (Chem.Mol) – The molecule to get the adjacency matrix of.
Returns:: npt.NDArray[np.int_] – A square adjacency matrix of the molecule, where a 1 indicates that atoms are bonded and a 0 indicates that atoms aren’t bonded.

rdtools.dist.get_bond_distance_matrix(mol: Mol) → ndarray[tuple[int, ...], dtype[float64]]#

Get the bond distance matrix of the molecule.

Parameters:: mol (Chem.Mol) – The molecule to get the bond distance matrix of.
Returns:: npt.NDArray[np.float64] – A square bond distance matrix of the molecule, where the value of each element is the bond distance between the atoms.

rdtools.dist.get_colliding_atoms(mol: Mol, conf_id: int = -1, threshold: float = 0.4, reference: Literal['vdw', 'covalent', 'bond'] = 'vdw') → list[tuple[int, int]]#

Get the atom pairs that are considered colliding.

The atoms are considered to be colliding if the distance between two atoms <= threshold * reference distance. The reference distance is based on the van der Waals radius, covalent radius or bond radius.

Parameters:

mol (Chem.Mol) – The molecule to check for colliding atoms.
conf_id (int, optional) – The conformer ID of the molecule to get the distance matrix of. Defaults to -1.
threshold (float, optional) – A multiplier applied on the reference matrix . Defaults to 0.4.
reference (Literal["vdw", "covalent", "bond"], optional) – The reference matrix to use. Defaults to "vdw". Options: - "vdw" for reference distance based on van de Waals radius - "covalent" for reference distance based on covalent radius - "bond" for reference distance based on bond radius

Returns:

list[tuple[int, int]] – A list of tuples of the atom indices that are potentially colliding.

rdtools.dist.get_covalent_distance_matrix(mol: Mol) → ndarray[tuple[int, ...], dtype[float64]]#

Generate a covalent matrix of the molecule.

Parameters:: mol (Chem.Mol) – The molecule to generate the derived covalent matrix of.
Returns:: npt.NDArray[np.float64] – A square covalent matrix of the molecule, where the value of each element is the covalent radius of the atoms.

rdtools.dist.get_distance_matrix(mol: Mol, conf_id: int = -1, balaban: bool = False) → ndarray[tuple[int, ...], dtype[float64]]#

Get the distance matrix of the molecule.

Parameters:

mol (Chem.Mol) – The molecule to get the distance matrix of.
conf_id (int, optional) – The conformer ID of the molecule to get the distance matrix of. Defaults to -1.
balaban (bool, optional) – Whether to use the Balaban distance. Defaults to False. If True, the distance matrix will be calculated using the Balaban distance.

Returns:

npt.NDArray[np.float64] – A square distance matrix of the molecule, where the value of each element is the distance between the atoms.

rdtools.dist.get_missing_bonds(mol: Mol, conf_id: int = -1, threshold: float = 1.5, reference: Literal['vdw', 'covalent', 'bond'] = 'covalent') → list[tuple[int, int]]#

Check whether the molecule has missing bonds heuristically.

If the distance between two atoms <= threshold * reference distance, the bond between the atoms are considered to be missing.

Parameters:

mol (Chem.Mol) – The molecule to check for missing bonds.
conf_id (int, optional) – The conformer ID of the molecule to get the distance matrix of. Defaults to -1.
threshold (float, optional) – A multiplier applied on the reference matrix . Defaults to 1.5.
reference (Literal["vdw", "covalent", "bond"], optional) – The reference matrix to use. Defaults to "covalent". Options: - "vdw" for reference distance based on van de Waals radius - "covalent" for reference distance based on covalent radius - "bond" for reference distance based on bond radius

Returns:

list[tuple[int, int]] – A list of tuples of the atom indices that are potentially missing bonds.

rdtools.dist.get_shortest_path(mol: Mol, idx1: int, idx2: int) → tuple[int, ...]#

Get the shortest path between two atoms in a molecule.

The RDKit GetShortestPath has a very long setup time ~ 0.5ms (on test machine) regardless of the size of the molecule. As a comparison, on the same machine, a naive python implementation of DFS (_find_shortest_path) takes ~0.5 ms for a 100-C normal alkane end to end. Therefore, it make more sense to use a method with a shorter setup time though scaling worse for smaller molecules while using GetShortestPath for larger molecules.

Parameters:

mol (Chem.Mol) – The molecule to be checked.
idx1 (int) – The index of the first atom.
idx2 (int) – The index of the second atom.

Returns:

tuple[int, …] – A list of atoms in the shortest path between the two atoms.

rdtools.dist.get_vdw_distance_matrix(mol: Mol) → ndarray[tuple[int, ...], dtype[float64]]#

Generate a van der Waals matrix of the molecule.

Parameters:: mol (Chem.Mol) – The molecule to generate the derived van der Waals matrix of.
Returns:: npt.NDArray[np.float64] – A square van der Waals matrix of the molecule, where the value of each element is the van der Waals radius of the atoms.

rdtools.dist.has_colliding_atoms(mol: Mol, conf_id: int = -1, threshold: float = 0.4, reference: Literal['vdw', 'covalent', 'bond'] = 'vdw') → bool#

Check whether the molecule has colliding atoms.

If the distance between two atoms <= threshold * reference distance, the atoms are considered to be colliding.

Parameters:

mol (Chem.Mol) – The molecule to check for colliding atoms.
conf_id (int, optional) – The conformer ID of the molecule to get the distance matrix of. Defaults to -1.
threshold (float, optional) – A multiplier applied on the reference matrix . Defaults to 0.4.
reference (Literal["vdw", "covalent", "bond"], optional) – The reference matrix to use. Defaults to "vdw". Options: - "vdw" for reference distance based on van de Waals radius - "covalent" for reference distance based on covalent radius - "bond" for reference distance based on bond radius

Returns:

bool – Whether the molecule has colliding atoms.