Peter,

adptbx.u_as_b(-1) = -78.96.

When you read in a model with ANISOU records the equivalent isotropic B is ignored but instead the anisotropic one is read in. After you did your manipulations you write out a PDB file with isotropic ADP which results in -78.96. I quickly looked at your code and it seems to me  that you don't take this into account. Please use or look for an example in  model.py facility to write out a PDB file: this is the common place we use to write PDB file after phenix.refine. Please let me know if it needs to be adjusted for your particular task or feel free to do it yourself.

Pavel.


Peter Zwart wrote: 
NOt sure what happens.

I have to investigate.
Could you send me example data and model? I'll have a look and see
where the problem lies.

Thanks

Peter

2007/6/13, Kate Kavanagh <[email protected]>:
  
Hi Peter,

The reindexing works fine but when I try to convert back to R3:H all atoms with ANISOU records are given B-factors of -79.86:

ATOM      1  N   MET A   1      16.402  60.069  44.768  1.00-78.96           N
ATOM      2  CA  MET A   1      16.316  58.621  44.978  1.00-78.96           C
ATOM      3  CB  MET A   1      17.622  57.922  44.573  1.00-78.96           C
ATOM      4  C   MET A   1      15.966  58.300  46.432  1.00-78.96           C
ATOM      5  O   MET A   1      16.196  59.115  47.326  1.00-78.96           O
ATOM      6  N   PHE A   2      15.408  57.114  46.661  1.00-78.96           N
ATOM      7  CA  PHE A   2      15.047  56.686  48.013  1.00-78.96           C
ATOM      8  CB  PHE A   2      13.750  55.866  48.016  1.00-78.96           C
ATOM      9  CG  PHE A   2      13.323  55.436  49.391  1.00-78.96           C
.....

And since TLS refinement has been used, all protein atoms have ANISOU records.

Any suggestions?

Kate

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

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