def _from_pdb_text(cls, pdb_text, verbose=False):
"""
Get structural data from pdb text as DataFrame.
Parameters
----------
pdb_text : str
Pdb file content from KLIFS database.
verbose : bool
Show only default columns (False) or additionally input-format specific columns (True).
Returns
-------
dict of pandas.DataFrame
Structural data
"""
# Set pdb columns (index, name, dtype) as DataFrame
pdb_columns = pd.DataFrame.from_dict(PDB_COLUMNS,
orient="index",
columns=["name", "dtype"])
# Use biopandas to parse the pdb format and return DataFrames
# TODO in the future: BioPandas: wait for pdb equivalent of PandasMol2.read_mol2_from_list
ppdb = PandasPdb()
pdb_dict = ppdb._construct_df(pdb_text.splitlines(True))
# Concatenate ATOM and HETATM entries
pdb_df = pd.concat([pdb_dict["ATOM"],
pdb_dict["HETATM"]]).reset_index(drop=True)
# Select only columns of interest and rename columns
pdb_df = pdb_df.iloc[:, pdb_columns.index.to_list()]
pdb_df.columns = pdb_columns["name"].to_list()
# Merge residue ID and insertion code
pdb_df["residue.id"] = pdb_df.apply(
lambda x: str(x["residue.id"]) + x["residue.insertion"], axis=1)
# Format DataFrame
pdb_df = cls._format_dataframe(pdb_df, verbose)
if len(pdb_df) == 0:
raise ValueError(
f"No structural data could be loaded. Is the input text in pdb format?"
)
return pdb_df
def test__construct_df():
"""Test pandas dataframe construction"""
ppdb = PandasPdb()
dfs = ppdb._construct_df(three_eiy.splitlines())
assert set(dfs.keys()) == {'OTHERS', 'ATOM', 'ANISOU', 'HETATM'}
assert set(dfs['ATOM'].columns) == set(ATOM_DF_COLUMNS)
assert set(dfs['HETATM'].columns) == set(ATOM_DF_COLUMNS)
assert set(dfs['ANISOU'].columns) == set(ANISOU_DF_COLUMNS)
exp = pd.Series(np.array(['ATOM', 1, '', 'N', '', 'SER',
'', 'A', 2, '', '', 2.527, 54.656, -1.667, 1.0,
52.73, '', '', 'N', None, 609]),
index=['record_name', 'atom_number', 'blank_1',
'atom_name', 'alt_loc', 'residue_name',
'blank_2', 'chain_id', 'residue_number',
'insertion', 'blank_3',
'x_coord', 'y_coord', 'z_coord',
'occupancy', 'b_factor', 'blank_4',
'segment_id', 'element_symbol',
'charge', 'line_idx'])
assert exp.equals(dfs['ATOM'].loc[0, :])
def _pdb_text_to_dataframe(pdb_text):
"""
Get structural data from pdb text.
Parameters
----------
pdb_text : str
Pdb file content from KLIFS database.
Returns
-------
dict of pandas.DataFrame
Structural data
"""
ppdb = PandasPdb()
pdb_dict = ppdb._construct_df(pdb_text.splitlines(True))
print(f'Structural data keys: {pdb_dict.keys()}')
return pdb_dict
