Convert mmCIF file to pandas.Dataframe.

cif2dataframe.py

import gemmi
import pandas as pd

cif_doc_block = gemmi.cif.read('1ug6.cif.gz')[0]

keys = (
    '_atom_site.id',
    '_atom_site.label_atom_id',
    '_atom_site.label_comp_id',
    '_atom_site.label_asym_id',
    '_atom_site.auth_asym_id',
    '_atom_site.label_entity_id',
    '_atom_site.label_seq_id',
    '_atom_site.auth_seq_id',
    '_atom_site.pdbx_PDB_ins_code',
    '_atom_site.Cartn_x',
    '_atom_site.Cartn_y',
    '_atom_site.Cartn_z',
    '_atom_site.B_iso_or_equiv',
    '_atom_site.pdbx_PDB_model_num',
)

pdb_frame = pd.DataFrame({key: cif_doc_block.find_values(key) for key in keys})
for key in keys[9:-1]:
    pdb_frame[key] = pdb_frame[key].astype('float32')
for key in (keys[0], keys[-1]):
    pdb_frame[key] = pdb_frame[key].astype('int')