Molecules

Molecules#

Reading RDKit molecules — prolif.rdkitmol#

class prolif.rdkitmol.BaseRDKitMol[source]#

Bases: Mol

Base molecular class that behaves like an RDKit Mol with extra attributes (see below). The sole purpose of this class is to define the common API between the Molecule and Residue classes. This class should not be instantiated by users.

Parameters:

mol (rdkit.Chem.rdchem.Mol) – A molecule (protein, ligand, or residue) with a single conformer

centroid#

XYZ coordinates of the centroid of the molecule

Type:

rdkit.Geometry.rdGeometry.Point3D

xyz#

XYZ coordinates of all atoms in the molecule

Type:

numpy.ndarray

Reading proteins and ligands — prolif.molecule#

class prolif.molecule.Molecule(mol: Mol, *, use_segid: bool = False)[source]#

Bases: BaseRDKitMol

Main molecule class that behaves like an RDKit Mol with extra attributes (see examples below). The main purpose of this class is to access residues as fragments of the molecule.

Parameters:

  • mol (rdkit.Chem.rdchem.Mol) – A ligand or protein with a single conformer

  • use_segid (bool, default = False) – Use the segment number rather than the chain identifier as a chain.

residues#

A dictionnary storing one/many Residue indexed by ResidueId. The residue list is sorted.

Type:

prolif.residue.ResidueGroup

n_residues#

Number of residues

Type:

int

Examples

In [1]: import MDAnalysis as mda

In [2]: import prolif

In [3]: u = mda.Universe(prolif.datafiles.TOP, prolif.datafiles.TRAJ)

In [4]: mol = u.select_atoms("protein").convert_to("RDKIT")

In [5]: mol = prolif.Molecule(mol)

In [6]: mol
Out[6]: <prolif.molecule.Molecule with 302 residues and 4988 atoms at 0x7f2d1ea93f60>

You can also create a Molecule directly from a Universe:

In [7]: mol = prolif.Molecule.from_mda(u, "protein")

In [8]: mol
Out[8]: <prolif.molecule.Molecule with 302 residues and 4988 atoms at 0x7f2d1f7cf290>

Notes

Residues can be accessed easily in different ways:

In [9]: mol["TYR38.A"] # by resid string (residue name + number + chain)
Out[9]: <prolif.residue.Residue TYR38.A at 0x7f2d1eaf1d50>

In [10]: mol[42] # by index (from 0 to n_residues-1)
Out[10]: <prolif.residue.Residue LEU80.A at 0x7f2d1ea9e7a0>

In [11]: mol[prolif.ResidueId("TYR", 38, "A")] # by ResidueId
Out[11]: <prolif.residue.Residue TYR38.A at 0x7f2d1eaf1d50>

See prolif.residue for more information on residues

Changed in version 2.1.0: Added use_segid.

classmethod from_mda(obj: MDAObject, selection: str | None = None, *, use_segid: bool | None = None, **kwargs: Any) → Molecule[source]#

Creates a Molecule from an MDAnalysis object

Parameters:

Example

In [1]: mol = prolif.Molecule.from_mda(u, "protein")

In [2]: mol
Out[2]: <prolif.molecule.Molecule with 302 residues and 4988 atoms at 0x7f2d1eac5670>

Which is equivalent to:

In [3]: protein = u.select_atoms("protein")

In [4]: mol = prolif.Molecule.from_mda(protein)

In [5]: mol
Out[5]: <prolif.molecule.Molecule with 302 residues and 4988 atoms at 0x7f2d1e92ca90>

Changed in version 2.1.0: Added use_segid.

classmethod from_rdkit(mol: Mol, resname: str = 'UNL', resnumber: int = 1, chain: str = '', use_segid: bool = False) → Molecule[source]#

Creates a Molecule from an RDKit molecule

While directly instantiating a molecule with prolif.Molecule(mol) would also work, this method insures that every atom is linked to an AtomPDBResidueInfo which is required by ProLIF

Parameters:

  • mol (rdkit.Chem.rdchem.Mol) – The input RDKit molecule

  • resname (str) – The default residue name that is used if none was found

  • resnumber (int) – The default residue number that is used if none was found

  • chain (str) – The default chain Id that is used if none was found

  • use_segid (bool, default = False) – Use the segment number rather than the chain identifier as a chain

Notes

This method only checks for an existing AtomPDBResidueInfo in the first atom. If none was found, it will patch all atoms with the one created from the method’s arguments (resname, resnumber, chain).

Changed in version 2.1.0: Added use_segid.

class prolif.molecule.mol2_supplier(path: Union[str, Path], cleanup_substructures: bool = True, sanitize: bool = True, **kwargs: Any)[source]#

Bases: Sequence[Molecule]

Supplies molecules, given a path to a MOL2 file

Parameters:

  • path (str) – A path to the .mol2 file

  • sanitize (bool) – Whether to sanitize each molecule or not.

  • cleanup_substructures (bool) – Toggles standardizing some substructures found in mol2 files, based on atom types.

  • resname (str) – Residue name for every ligand

  • resnumber (int) – Residue number for every ligand

  • chain (str) – Chain ID for every ligand

Returns:

suppl – A sequence that provides Molecule objects

Return type:

Sequence

Example

The supplier is typically used like this:

>>> lig_suppl = mol2_supplier("docking/output.mol2")
>>> for lig in lig_suppl:
...     # do something with each ligand

Changed in version 1.0.0: Molecule suppliers are now sequences that can be reused, indexed, and can return their length, instead of single-use generators.

Changed in version 2.1.0: Added cleanup_substructures and sanitize parameters (default to True, same behavior as before).

class prolif.molecule.pdbqt_supplier(paths: Iterable[Union[str, Path]], template: Mol, converter_kwargs: dict | None = None, **kwargs: Any)[source]#

Bases: Sequence[Molecule]

Supplies molecules, given paths to PDBQT files

Parameters:

  • paths (list) – A list (or any iterable) of PDBQT files

  • template (rdkit.Chem.rdchem.Mol) – A template molecule with the correct bond orders and charges. It must match exactly the molecule inside the PDBQT file.

  • converter_kwargs (dict | None) – Keyword arguments passed to the RDKitConverter of MDAnalysis

  • resname (str) – Residue name for every ligand

  • resnumber (int) – Residue number for every ligand

  • chain (str) – Chain ID for every ligand

Returns:

suppl – A sequence that provides Molecule objects

Return type:

Sequence

Example

The supplier is typically used like this:

>>> import glob
>>> pdbqts = glob.glob("docking/ligand1/*.pdbqt")
>>> lig_suppl = pdbqt_supplier(pdbqts, template)
>>> for lig in lig_suppl:
...     # do something with each ligand

Changed in version 1.0.0: Molecule suppliers are now sequences that can be reused, indexed, and can return their length, instead of single-use generators.

Changed in version 1.1.0: Because the PDBQT supplier needs to strip hydrogen atoms before assigning bond orders from the template, it used to replace them entirely with hydrogens containing new coordinates. It now directly uses the hydrogen atoms present in the file and won’t add explicit ones anymore, to prevent the fingerprint from detecting hydrogen bonds with “random” hydrogen atoms. A lot of irrelevant warnings and logs have been disabled as well.

class prolif.molecule.sdf_supplier(path: str, sanitize: bool = True, **kwargs: Any)[source]#

Bases: Sequence[Molecule]

Supplies molecules, given a path to an SDFile

Parameters:

  • path (str) – A path to the .sdf file

  • sanitize (bool) – Whether to sanitize each molecule or not.

  • resname (str) – Residue name for every ligand

  • resnumber (int) – Residue number for every ligand

  • chain (str) – Chain ID for every ligand

Returns:

suppl – A sequence that provides Molecule objects. Can be indexed

Return type:

Sequence

Example

The supplier is typically used like this:

>>> lig_suppl = sdf_supplier("docking/output.sdf")
>>> for lig in lig_suppl:
...     # do something with each ligand

Changed in version 1.0.0: Molecule suppliers are now sequences that can be reused, indexed, and can return their length, instead of single-use generators.

Changed in version 2.1.0: Added sanitize parameter (defaults to True, same behavior as before).

class prolif.residue.Residue(mol: Chem.Mol, *, use_segid: bool = False)[source]#

Bases: BaseRDKitMol

A class for residues as RDKit molecules

Parameters:

  • mol (rdkit.Chem.rdchem.Mol) – The residue as an RDKit molecule

  • use_segid (bool, default = False) – Use the segment number rather than the chain identifier as a chain

resid#

The residue identifier

Type:

prolif.residue.ResidueId

Notes

The name of the residue can be converted to a string by using str(Residue)

Changed in version 2.1.0: Added use_segid.

Contents