Search code examples
pythonpython-modulebiopython

Why would Bio.PDB.MMCIF2Dict() be missing?

I'm trying to see how much "header" info I can get out of a PDB file, following the hints at The Biopython Structural Bioinformatics FAQ. I'm led to use the mmCIF version. but there seems something odd about the package/egg:

>>> from Bio.PDB import *
>>> cifFile = '.../path2/3VQ8.cif'
>>> mmcif_dict = MMCIF2Dict(cifFile)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
NameError: name 'MMCIF2Dict' is not defined

I get a different error if I fully qualify the class:

>>> import Bio.PDB
>>> mmcif_dict = Bio.PDB.MMCIF2Dict(cifFile)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'MMCIF2Dict'

Note that most of Bio.PDB works as I expect:

>>> parser=PDBParser(PERMISSIVE=1)
>>> inf = '.../path2/3NF8.pdb'
>>> structure=parser.get_structure('3Nf8',inf)
>>> structure.header.keys()
['structure_method', 'head', 'journal_reference', 'compound',
 'name', 'author', 'deposition_date', 'release_date', 'source',
 'resolution', 'structure_reference']

and I have a recent version:

>>> Bio.__version__
'1.61'

which does have the MMCIF2Dict.py file is inside the Bio egg, but MMCIF2Dict isn't in the module:

>>> dir(Bio.PDB)
['AbstractPropertyMap', 'Atom', 'CaPPBuilder', 'Chain', 'DSSP', 'Dice',
 'Entity', 'ExposureCN', 'FragmentMapper', 'HSExposure', 'HSExposureCA',
 'HSExposureCB', 'Model', 'NeighborSearch', 'PDBExceptions', 'PDBIO',
 'PDBList', 'PDBParser', 'PPBuilder', 'Polypeptide', 'Residue',
 'ResidueDepth', 'Select', 'Selection', 'Structure',
 'StructureAlignment', 'StructureBuilder', 'Superimposer', 'Vector',
 '__builtins__', '__doc__', '__file__', '__name__', '__package__',
 '__path__', 'calc_angle', 'calc_dihedral', 'extract', 'get_surface',
 'is_aa', 'm2rotaxis', 'make_dssp_dict', 'mmCIF', 'parse_pdb_header',
 'refmat', 'rotaxis', 'rotaxis2m', 'rotmat', 'standard_aa_names',
 'to_one_letter_code', 'vector_to_axis']

Anyone have a clue?

Solution

  • You'll find the answer in Bio.PDB's __init__.py.

    In short, from Bio.PDB import * imports everything in __init__.py's namespace. If you read __init__.py, you can see that it imports MMCIFParser, but not MMCIF2Dict (at least not directly): MMCIFParser imports MMCIF2Dict and makes use of it there.

    So, if you do from Bio.PDB import MMCIF2Dict; mmcif_dict = MMCIF2Dict.MMCIF2Dict(cifFile), it should work without complaint. However, from Bio.PDB import * won't put MMCIF2Dict into the current namespace.