[phenixbb] Weird Hydorgen refinement

Pavel Afonine pafonine at lbl.gov
Mon Aug 22 18:22:03 PDT 2011

Hi Yuri,

> The pdb did have anisou records for H's. I believe that is because in 
> some previous runs I had only selected chain 'A' - I left out "and not 
> elemnt H" - which I believe caused the H's to get refined 
> anisotropically.

Ok, that would be the explanation.

> I am trying to get rid of the ANISOU records and Ill run refinement 
> again.


> ps. I am getting ready to deposit the coordinates and I dont want to 
> keep hydrogens in the model, how can I get rid of them? I ve been 
> using pyMol to remove them, but pymol gets rid of all the info 
> -including symmetry -that phenix adds...

Any post-refinement changes to the model are very bad. By doing so you 
invalidate the reported in REMARK 3 statistics, such as R-factors, atom 
counts, etc. This is explained in details here:

   phenix.model_vs_data: a high-level tool for the calculation of
   crystallographic model and data statistics. P.V. Afonine, R.W.
   Grosse-Kunstleve, V.B. Chen, J.J. Headd, N.W. Moriarty, J.S. Richardson,
   D.C. Richardson, A. Urzhumtsev, P.H. Zwart, P.D. Adams
   J. Appl. Cryst. 43, 677-685 (2010).

In general, phenix.pdbtools is the tool to manipulate PDB files:

phenix.pdbtools model.pdb remove="element H"

will remove all H atoms. If you do so then re-run the refinement so you 
get correct numbers.

More examples of PDB tools usage:


More information about the phenixbb mailing list