Zero or low B-factor in pdb file

Model Building
1

Hi all,

I am seeking advice regarding B-factors in a PDB file. I have completed model building, but when I open the PDB file, I notice that many atoms have either zero or unusually low B-factors which I believe should not be the case. I tried running real-space refinement in phenix after replacing zero B-factors with random values, but it didn’t work. I would appreciate any insights or suggestions on how to address this.

Thanks in advance

2

Hi Wynn,

We are having the same issues here with Phenix refinement - did you get any help with this?
Perhaps I should be over on the phenix bb; we’ve tried Phenix v 1.20.1-4487 and 1.21.2-5419, both giving B-factor 0 for e.g. side chain Carbons and Calphas in good density.

Hope we can find a fix!

3

Hi Fiona,

My colleague and I had the same issue. In his case, it was resolved by re-running the refinement using the newest version available at the time. In my case, the ligand wasn’t being recognized (it was an amino acid derivative), and once that was fixed, the B-factors returned to normal.
Sorry I don’t have a more definitive solution, but I hope this helps in some way.

4

Thanks Wynn, we’ll try updating and see if that fixes the issue. We have noted some cryoEM structures deposited in the pdb refined with Phenix that have zero B-factors for some atoms, so perhaps a broader issue.

5

I had the same situation. After I reset the “Nproc” from 8 to 1, the problem was solved. There is a FAQ from the phenix website " I set the number of processors to 8 - why is phenix.refine still only using a single processor?

As noted above, the standard parallelization primarily affects weight optimization. If this is not performed, most of the program will run serially."
I guess that may be the reason why we should use 1 processor for real space refinement.
Best,
Jian