Hi Xiao Lei,

(1) I assume that Refmac does not allow same ID, residue number and atom name, regardless of occupancy (I am just guessing from the output).
Usually, alternate conformations require an alternate location indicator (column 17 in the pdb file). As far as I understand your description and the log, there are no altloc IDs in your input file. So the program does not understand that it is meant as alternate conformation.

Did you get a similar error message when you used the files before modification with phenix tools, i.e. did Refmac also show an error before you changed the occupancy?
If the error only occurred with phenix modified file, could you please send the pdb files before and after modification, so we can check why the file produces an error? Please send the files off-list (answer only to me), they will be of course treated confidentially.

(2) It looks like it is as you said: Refmac lists possible links because some atoms are close together. But according to the output, they were not used and possibly did not have any effect on your refinement - but difficult to say without seeing the result.

(3) If you want to use phenix.refine, please have a look at the Computational Crystallography Newsletter article about occupancy refinement. Your case might be in there:
https://www.phenix-online.org/newsletter/CCN_2015_07.pdf#page=12

Best wishes,

Dorothee

On Fri, Apr 29, 2016 at 1:42 PM, Xiao Lei <xiaoleiusc@gmail.com> wrote:
Hi Phenixbb members,

I'm using Phenix PDB File Editor under Phenix GUI "Model Tool" to change occupancy of two short 7 polyalanine alpha helix peptide chain (chain I and K, placed in reverse orientation of each other in one region (or density blob) of my model, each chain half occupancy) , I can change the occupancy very easily with the tool, but I have two types of errors using the coordinates generated by the tool in Refmac. 

1. If I set chain ID of chain K to chain I with the same residue number with half occupancy but in reverse order (Chain I, Ala7th--13th 0.5 occupancy and Chain I, Ala7th--13th 0.5 occupancy) , Refmac could not refine the model with an error message below:
 I am reading library. Please wait.
                mon_lib.cif
  ERROR : ALA              : duplicated atom_name : "N   ".
          chain: II   residue:    7
  ERROR : ALA              : duplicated atom_name : "CA  ".
          chain: II   residue:    7
  ERROR : ALA              : duplicated atom_name : "CB  ".
          chain: II   residue:    7
  ERROR : ALA              : duplicated atom_name : "C   ".
          chain: II   residue:    7
  ERROR : ALA              : duplicated atom_name : "O   ".
          chain: II   residue:    7
 IERR =            1

There is an error. See above
===> Error: Fatal error. Cannot continue.

I understand Refmac can not allow two peptide chain with the same ID and residue number if both peptides are full occupancy (1), but why in this case even half occupancy of the same chain ID and residue number is not tolerated?

2. If I set chain ID of chain K and chain I half occupancy (Chain K, Ala7th--13th 0.5 occupancy and chain I, Ala7th--13th 0.5 occupancy) , put chain K and chain I in the same density blob with reverse direction, and run Refmac, this time Refmac can finish but gave me error message in log file as below:
  I am reading library. Please wait.

                mon_lib.cif

  INFO:     link is found (not be used) dist=   1.486 ideal_dist=   1.240

            ch:II   res:   7  ALA              at:N   .->ch:KK   res:  11  ALA              at:O   .

  INFO:     link is found (not be used) dist=   1.625 ideal_dist=   1.080

            ch:II   res:   8  ALA              at:O   .->ch:KK   res:   9  ALA              at:O   .

  INFO:     link is found (not be used) dist=   1.584 ideal_dist=   1.400

            ch:II   res:  10  ALA              at:N   .->ch:KK   res:  11  ALA              at:N   .

  INFO:     link is found (not be used) dist=   1.397 ideal_dist=   1.530

            ch:II   res:  10  ALA              at:CB  .->ch:KK   res:  11  ALA              at:CB  .

  INFO:     link is found (not be used) dist=   1.153 ideal_dist=   1.400

            ch:II   res:  11  ALA              at:N   .->ch:KK   res:  10  ALA              at:N   .

  INFO:     link is found (not be used) dist=   1.820 ideal_dist=   1.524

            ch:II   res:  11  ALA              at:CA  .->ch:KK   res:  10  ALA              at:CB  .

  INFO:     link is found (not be used) dist=   0.888 ideal_dist=   1.530

            ch:II   res:  11  ALA              at:CB  .->ch:KK   res:  10  ALA              at:CB  .

  INFO:     link is found (not be used) dist=   1.808 ideal_dist=   1.530

            ch:II   res:  12  ALA              at:CB  .->ch:KK   res:   9  ALA              at:CB  .

  INFO:     link is found (not be used) dist=   1.764 ideal_dist=   1.400

            ch:II   res:  13  ALA              at:N   .->ch:KK   res:   7  ALA              at:N   .

  --------------------------------
I do not know why Refmac gave me this message, is it because the two helix chain (I and K) are placed in the same region, even though both with 0.5 occupancy of both chain, Refmac still automatically tries to connect the main chain atom between two different chains (I and K) because main chain atoms are placed too close to each other (Refmac think there is a link between them?) ?


3. In summary, should I use Phenix Refine instead of Refmac to handle the pdb coordinates generated by Phenix PDB file editor because Phenix Refine may have better compatibility?

Thanks in advance.


_______________________________________________
phenixbb mailing list
phenixbb@phenix-online.org
http://phenix-online.org/mailman/listinfo/phenixbb
Unsubscribe: phenixbb-leave@phenix-online.org