KeyError: 'B_iso_or_equiv'

[frgoe003]

Hello, I'm trying to process a CIF file with Prody, but encounter a KeyError: 'B_iso_or_equiv'. Not sure what happens here, as the header structure is the same as in other cases which can be parsed.

Attaching the example file:
1ips_B.cif.zip

  File "/data/chem-cat/fgoerlich/miniconda3/envs/docking/lib/python3.11/site-packages/prody/proteins/ciffile.py", line 122, in parseMMCIF
    result = parseMMCIFStream(cif, chain=chain, **kwargs)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/data/chem-cat/fgoerlich/miniconda3/envs/docking/lib/python3.11/site-packages/prody/proteins/ciffile.py", line 197, in parseMMCIFStream
    _parseMMCIFLines(ag, lines, model, chain, subset, altloc)
  File "/data/chem-cat/fgoerlich/miniconda3/envs/docking/lib/python3.11/site-packages/prody/proteins/ciffile.py", line 427, in _parseMMCIFLines
    bfactors[acount] = line.split()[fields['B_iso_or_equiv']]
                                    ~~~~~~^^^^^^^^^^^^^^^^^^
KeyError: 'B_iso_or_equiv'

[jamesmkrieger]

MmCIF files can have peculiarities. It could be that they don’t have this field or got it wrong somehow.

We’ll have a look and let you know when we have a fix or suggestion

[frgoe003]

Okay it seems that without the B_iso_or_equiv entry the cif file can't be parsed in general, so not an error of ProDy. Why is that the case?

[jamesmkrieger]

We could still make prody able to parse files that lack that field though. Why is the file lacking that field?

[frgoe003]

I've used Biotite to generate the CIF file. If you don't explicitly request b_factor when using their get_structure method, it misses that field. I'm using the CIF files to generate PDBQT files for docking via Meeko (which uses Prody), not sure if the B_iso_or_equiv information is necessary for that.

[jamesmkrieger]

I’d imagine it isn’t but I’ve no idea

[frgoe003]

Well smth for another day. Thanks anyway for the quick reply!

[jamesmkrieger]

You’re welcome. Yes, something for another day

[jamesmkrieger]

I’d say we should keep it open. If there’s a program that writes mmCIF without b factors then prody needs to be able to read them