Dear All, I´m refining a 2.1A with phenix 1.5-2 and get when I click on automatically add hydrogens to model following error message: refinement.pdb_interpretation.clash_guard.failure: Number of nonbonded interaction distances < 0.5:391 I suppose that this comes, because coot mixes up the hydrogens. See Re: [COOT] hydrogen restraints Paul Emsley Wed, 09 Dec 2009 14:11:34 -0800 Sara Zueger wrote: Hi everybody, Hi Sara, as I did not find a final solution to the hydrogen restraints problem (flying or exploding during real space refinement) in any of the mailing lists, I wanted to pick this topic up again. If one adds hydrogens by phenix.ready_set or reduce the labels for these hydrogens on let's say a Cbeta atom will be HB2 and HB3 (according to the pdb v3.2 convention). However, the coot/ccp4 library names them HB1 and HB2, so there are no restraints for HB3 atoms. I see that probably this naming problem can be overcome by using phenix.reduce -OLD, or renaming them manually in the pdb file (but if one deposits the structure to PDB does one have to name them back again?). Is there already some new solution? There is not a solution yet. And how about hydrogen restraints for the most N-terminal amino acid, where one has H1, H2 and H3? How are these handled? There seem to be no restraints for those as well, because they also fly around during real space refinement. Same situation. We need to do some drudge work to convert the hydrogen names in the libraries. AFAICS it is straightforward, but just hasn't been done by anyone in the CCP4/Coot community. I wonder what dictionary the Phenix fellow use? If they have updated the dictionary to 3.2 format, perhaps we could use their copy too. (However, we deal with this, it should be done by the end of the 0.7.x series.) Paul. What can - should I do now, to get my structure in the pdb, because just to remove them is according to the manual not allowed See: To remove hydrogens from a model: % phenix.pdbtools model.pdb remove="element H or element D" or Reduce programs can be used for this: % phenix.reduce model_h.pdb -trim > model_noH.pdb We strongly recommend to not remove hydrogen atoms after refinement since it will make the refinement statistics (R-factors, etc...) unreproducible without repeating exactly the same refinement protocol. 2. Have you ever seen that coot looses the restraints? I have a heme in the structure which I could refine without any problems, but now he gives me following error message: No restraints found Non-existent or minmal description of restrained residues. Are you sure that you read a non-minimal mmCIF dictionary for this monomer? Are you sure the PDB residue name matches the dictionary residue name? If not, try File -Import CIF dictionary Alternatively, did you check that the atom names of the PDB file matches those of the restraints. Thank you very much for your help. Georg -----Ursprüngliche Nachricht----- Von: [email protected] [mailto:[email protected]] Im Auftrag von jonathan elegheert Gesendet: Dienstag, 19. Jänner 2010 18:32 An: PHENIX user mailing list Betreff: Re: [phenixbb] twin refinement in phenix-1.5-2 Hi Kelly, I encountered the same with later nightlies than 1.5-2 also. The problem is the "fix_rotamers = True " command. You have to return it to its default value "None". Apparently, some subroutines like that one don't work with twin_f_lsq algorithm. I believe because the fix_rotamers function depends on a map cc calculation which also in itself doesn't work with twin_f_lsq. At least that's my understanding. Jonathan Elegheert PhD student Ghent University Kelly Daughtry wrote:

Hello all, I am wondering if anyone has run across this problem and knows how to fix it.

I have a twinned dateset and I am using phenix.refine to refine the dataset. I have the full 1.5-2 installation.

When I include the twin law (-k, -h, -l) in the refinement, I get this error:

File "/usr/swr/phenix-1.5-2/cctbx_project/mmtbx/twinning/twin_f_model.py", line 1651, in map_coefficients "m_gradient" AssertionError

If I do not include the twin law, I do not get the error and the refinment continues as normal.

When I downgraded to Phenix.1.4-3 (full install), And use the same twin law, I am able to run phenix.refine without an error.

I have attached my definition file for phenix.1.5-2 amd the log file.

Thanks in advance for any help you can provide, Kelly Daughtry

******************************************************* Kelly Daughtry PhD Candidate Department of Physiology and Biophysics Boston University School of Medicine 590 Commonwealth Ave R 390 Boston MA, 02215 (P) 617-358-5548 ******************************************************* ------------------------------------------------------------------------

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

Hello,

I encountered the same with later nightlies than 1.5-2 also. The problem is the "fix_rotamers = True " command. You have to return it to its default value "None".

Apparently, some subroutines like that one don't work with twin_f_lsq algorithm. I believe because the fix_rotamers function depends on a map cc calculation which also in itself doesn't work with twin_f_lsq.

This is correct and sorry for the problem. I have to add a meaningful error message. Pavel.