Hi Wolfram,

The build_crystal_structures() method is aimed at extracting a cctbx xray.structure object from a small molecule cif file, and isn't aimed at mmcif-format files, so in your case it looks through all the data blocks in the provided CIF, can't find any suitable xray.structures, so returns an empty dictionary. If you simply want to access the raw CIF items from the file, then this code snippet shows how you can extract (e.g.) the Rwork from the 1NXC cif block:

import iotbx.cif

reader = iotbx.cif.reader(file_path="1NXC.cif")

cif = reader.model()

cif_block = cif["1NXC"]

# get a single item from cif_block

print cif_block["_refine.ls_R_factor_R_work"]

# get a looped item from cif_block

print list(cif_block["_citation_author.name"])

If you want to extract the model for your crystal structure, then probably the most useful thing to do is extract a pdb.hierarchy object (the cctbx's internal representation of a PDB/mmCIF structure file). This functionality can accept both PDB and mmCIF files as input:

import iotbx.pdb

# this works exactly the same for pdb files

pdb_input = iotbx.pdb.input(file_name="1NXC.cif")

pdb_hierarchy = pdb_input.construct_hierarchy()

Alternatively, if you are only reading mmCIF files, need access to both the raw cif items and the pdb_hierarchy, and are concerned about overhead of reading mmCIF twice (only really relevant for extremely large files), then you can pass the cif object created above to the iotbx.pdb.mmcif.cif_input�function:

import iotbx.pdb.mmcif

pdb_input = iotbx.pdb.mmcif.cif_input(cif_object=cif)

pdb_hierarchy = pdb_input.construct_hierarchy()

Please let me know if you have any more questions about using iotbx.cif!

Cheers,

Richard

On 6 November 2013 11:38, wtempel <wtempel@gmail.com> wrote:

Hello all,
please consider the scenario of a manuscript that, among other things, describes 20 crystal structures.
Suppose the journal editor requires a Table 1.
I now looking for a tool to prepare such table with high accuracy and little effort.

My PDB deposition workflow includes the preparation of mmcif coordinates with pdb_extract. The mmcif file is complete in the sense that after upload to RCSB-ADIT, virtually no additional information is required for deposition.

My intuition was to combine CSV and IOTBX python modules and produce the table via input of a list of mmcif files.
The current phenix table_1 tool provides some, but not all of that functionality, and I thought once familiar with the python modules, I would quickly write the script.

After confirming that the example at
http://cctbx.sourceforge.net/iotbx_cif/
works with my cctbx installation,
I attempted to transform that code to read in PDB mmcif coordinates:

<code>
test_structure = iotbx.cif.reader(
� file_path="/tmp/1NXC.cif").build_crystal_structures()["1NXC"]
</code>
which throws an exception:

<output>
Traceback (most recent call last):
� File "1NXC.py", line 3, in <module>
� � test_structure = iotbx.cif.reader(file_path="/tmp/1NXC.cif").build_crystal_structures()["1NXC"]

KeyError: '1NXC'
</output>
How do I need to write the code so that it returns a python object with access to all the cif items?
Thank you,
Wolfram Tempel

_______________________________________________
cctbxbb mailing list
cctbxbb@phenix-online.org
http://phenix-online.org/mailman/listinfo/cctbxbb