[phenixbb] DNA covalent bond angles deviate from PDB standard dictionary

Scott Classen sclassen at lbl.gov
Wed Jan 5 12:00:41 PST 2011


Hello,

I'm running into validation problems with DNA when validating on PDB website.

the validation report indicates that the refined value is significantly different that the PDBs standard dictionary.

Deviation  Residue  Chain  Sequence  AT1  -  AT2  -  AT3    Bond    Dictionary  Standard
             Name     ID     Number                          Angle      Value    Deviation
-------------------------------------------------------------------------------------------
     2.2     DG       D        6      N9   -  C1'  -  O4'    110.5     108.3        0.3
     2.6     DC       D        7      N1   -  C1'  -  O4'    110.9     108.3        0.3
     2.3     DC       D        8      N1   -  C1'  -  O4'    110.6     108.3        0.3
     3.0     DT       D        9      N1   -  C1'  -  O4'    111.3     108.3        0.3

I've looked through a few of these in the .geo file and it appears that phenix.refine is using different values.

for example,

angle pdb=" O4'  DG D   6 "
      pdb=" C1'  DG D   6 "
      pdb=" N9   DG D   6 "
    ideal   model   delta    sigma   weight residual
   108.40  110.18   -1.78 3.00e+00 1.11e-01 3.52e-01


ideal is specified as 108.4º so clearly my refinement has converged, but it has NOT converged on the RCSBs accepted value of 110.5º

What to do? Who is "right" phenix or RCSB?

Thanks,
Scott




More information about the phenixbb mailing list