CIF structure file format error

[yx0516]

When i use BioPython to read the CIF structure file predicted by Protenix. The following error happened:

However the CIF file of other AF3-like models can be read normally by BioPython.

Please give a hand about this.

Thanks a lot.

[yx0516]

@LanLanBoom @lark @guanjq

[cloverzizi]

Hi @yx0516 ,
we previously hadn't included the occupancy field in the CIF file, but we have just updated the code. You can either pull the latest version of the code to run the inference or resubmit the task on the WebServer.

[yx0516]

Hi @cloverzizi,

I pull the latest version of code and run the inference. And use the 'pip install .' to install the latest version and run the inference.
The upper error is fixed. However a new error happens:

Traceback (most recent call last):
File "/data/PRG/tools/apps/Protenix/run.py", line 814, in
get_results_single_run(args.topN)
File "/data/PRG/tools/apps/Protenix/run.py", line 671, in get_results_single_run
cif_to_pdb(cif_path,f"rank_{index+1}.pdb")
File "/data/PRG/tools/apps/Protenix/run.py", line 52, in cif_to_pdb
io.save(pdb)
File "/home/wemol/.local/lib/python3.10/site-packages/Bio/PDB/PDBIO.py", line 352, in save
raise PDBIOException(e)
Bio.PDB.PDBExceptions.PDBIOException: Chain id ('A0') exceeds PDB format limit.

The Chain name in CIF file is 'A0', i think it should be 'A'.
Please help to check this.
Thanks a lot.

[cloverzizi]

Hi @yx0516,
We updated the code so that in Protenix v0.4.1, the inferred output CIF's label_asym_id is changed from formats like 'A0' to just 'A'. This adjustment ensures that when converting to PDB format, as long as the number of chains does not exceed 26, the requirement for single-character Chain IDs is met.