This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Quality control metrics in mmCIF

I have downloaded from PDB 291 mmCIF files that have information for P53 structures. 289 of them have information for experimental strutures: 248 files correspond to X-RAY diffraction structures, 30 for NMR and 11 for EM. I would like to perform a quality control of the files, but I'm new to this format and I don't know which metrics I should check. Considering that the structures come from three different types of experiments, which metrics do you recommend checking in each type of experiment and is there any recommended threshold for them?

Thanks in advance

pdb protein mmcif

1 answer

There is a lot of useful information in the mmCIF file headers that can be relevant. Each method will have slightly different information available that will be relevant.

Here are a few of the more important factors I would take into consideration:

  1. Resolution (lower Å is better)
  2. R-work/free (quality and agreement of proposed models; x-ray only)
  3. Conformer agreement (RMSD; NMR)
  4. Clashscore (not in header; number of clashing atoms per 1000 atoms)

The list goes on, so it would be worth exploring the information in the headers and see what is relevant for your use case.

Log in to answer this question.