def from_file(cls, filepath, residue_ids, name="", residue_labels=None):
"""
Initialize Pocket object from structure protein file.
Attributes
----------
filepath : str or pathlib.Path
File path to structural protein data.
residue_ids : list of str
Pocket residue IDs.
name : str
Name of protein (default: empty string).
residue_labels : None or list of str
Pocket residue labels. Set to None by default.
Returns
-------
opencadd.structure.pocket.Pocket
Pocket object.
"""
pocket = cls()
pocket.name = name
pocket._filepath = filepath
pocket._data = DataFrame.from_file(filepath)
residue_ids, residue_labels = _format_residue_ids_and_labels(
residue_ids, residue_labels)
pocket._residue_ids = residue_ids
pocket._residue_labels = residue_labels
return pocket
def test_from_file(self, filepath, verbose):
"""
Test if input produces a DataFrame.
"""
dataframe = DataFrame.from_file(filepath, verbose)
isinstance(dataframe, pd.DataFrame)
def to_dataframe(self,
structure_klifs_id_or_filepath,
entity="complex",
extension="mol2"): # pylint: disable=W0221
filepath = self._to_filepath(structure_klifs_id_or_filepath, entity,
extension)
dataframe = DataFrame.from_file(filepath)
dataframe = self._add_residue_klifs_ids(dataframe, filepath)
return dataframe
def by_structure_klifs_id(self, structure_klifs_id, extension="mol2"): # pylint: disable=W0221
# Get kinase pocket from structure ID
structures_local = Structures(self._database,
self._path_to_klifs_download)
structure = structures_local.by_structure_klifs_id(
structure_klifs_id).squeeze()
# Get list of KLIFS positions (starting at 1) excluding gap positions
klifs_ids = [
index
for index, residue in enumerate(structure["structure.pocket"], 1)
if residue != "_"
]
# Load pocket coordinates from file
pocket_path = (self._path_to_klifs_download /
structure["structure.filepath"] / f"pocket.{extension}")
dataframe = DataFrame.from_file(pocket_path)
# Get number of atoms per residue
# Note: sort=False important otherwise negative residue IDs will be sorted to the top
number_of_atoms_per_residue = dataframe.groupby(
["residue.name", "residue.id"], sort=False).size()
# Get KLIFS position IDs for each atom in molecule
klifs_ids_per_atom = []
for klifs_id, n in zip(klifs_ids, number_of_atoms_per_residue):
klifs_ids_per_atom.extend([klifs_id] * n)
# Add column for KLIFS position IDs to molecule
dataframe["residue.klifs_id"] = klifs_ids_per_atom
dataframe = dataframe[["residue.id",
"residue.klifs_id"]].drop_duplicates()
# Add KLIFS IDs that are missing in pocket and fill with "_"
full_klifs_ids_df = pd.Series(range(1, 86),
name="residue.klifs_id").to_frame()
dataframe = full_klifs_ids_df.merge(dataframe,
on="residue.klifs_id",
how="left")
dataframe.fillna("_", inplace=True)
# Add column for KLIFS regions
dataframe = dataframe.merge(POCKET_KLIFS_REGIONS,
on="residue.klifs_id",
how="left")
dataframe = dataframe.astype({"residue.klifs_id": "Int64"})
# Standardize DataFrame
dataframe = self._standardize_dataframe(
dataframe,
DATAFRAME_COLUMNS["pockets"],
)
# Add KLIFS region and color TODO not so nice to have this after standardization
dataframe = self._add_klifs_region_details(dataframe)
return dataframe