Locating missing parts of the model

A difference map is calculated in order to locate either missing parts of the model or incorrectly placed atoms. The major error in the model at this point is the absence of several ytterbium ions which we know are bound in the calcium binding sites of the two molecules. We hope that these will be apparent in the difference map when contoured at positive sigma. Here a cross-validated, sigma-A weighted difference map is calculated using the CNS task file model_map_1fofc.inp:

      cns_solve < model_map_1fofc.inp > model_map_1fofc.out [30 seconds]

If you have mapman installed, you can use the command

      map_to_omap *.map

The map shows 4 outstanding peaks (all above 20 sigma). These are clearly separated from the other difference density peaks in the map. This can be seen both by visual inspection of the map but also in the positive density peak list generated (model_map_1fofc_positive.peaks). This peak list (which is in PDB format) is used to add the ytterbiums to the model.

The positive peak list:

ATOM   1749 PEAK PEAK    1      17.491  36.603  50.078  1.00 33.53      PEAK
ATOM   1750 PEAK PEAK    2      22.797  35.840  43.602  1.00 27.01      PEAK
ATOM   1751 PEAK PEAK    3       1.547  -5.134  15.074  1.00 25.92      PEAK
ATOM   1752 PEAK PEAK    4       8.461  -1.406  12.071  1.00 22.13      PEAK
ATOM   1753 PEAK PEAK    5      30.817  23.796  35.520  1.00  6.50      PEAK
ATOM   1754 PEAK PEAK    6      14.353  28.600  44.406  1.00  6.42      PEAK

The electron density map:

Sigma-A weighted difference map (at 7 sigma). Four large positive peaks are clearly seen.

In order to generate a new model with the ytterbium ions included we can merge the coordinates for the current model and the outstanding peaks from the peaks list. The atoms name, residue name and chain id for the ytterbium ions must be changed manually. The end of the new coordinate file look like this:

ATOM   1745  CG  PRO B 220      -4.556  16.898  32.163  1.00 48.84      B
ATOM   1746  C   PRO B 220      -3.260  19.366  30.884  1.00 58.22      B
ATOM   1747  O   PRO B 220      -3.209  19.542  32.133  1.00 58.22      B
ATOM   1748  OXT PRO B 220      -3.115  20.288  30.036  1.00 48.84      B
ATOM   1749 YB+3 YB3 Y   1      17.491  36.603  50.078  1.00 33.53
ATOM   1750 YB+3 YB3 Y   2      22.797  35.840  43.602  1.00 27.01
ATOM   1751 YB+3 YB3 Y   3       1.547  -5.134  15.074  1.00 25.92
ATOM   1752 YB+3 YB3 Y   4       8.461  -1.406  12.071  1.00 22.13
END

Note that the atom name has been changed to YB+3 and the residue name to YB3, these are the appropriate names for a ytterbium 3+ ion. This information can be obtained from the CNS topology file CNS_TOPPAR:ion.top. The chain identifier has been changed to Y. It is not necessary to include a segid because in the subsequent generate stage we will any look at the chainid. This coordinate file is then used to generate a new coordinate and molecular topology file for CNS. Here the CNS task file generate_easy_yb.inp is used:

      cns_solve < generate_easy_yb.inp > generate_easy_yb.out [2 seconds]

These new coordinates are then refined further. It is important to first obtain reasonable estimates for the ytterbium ion B-factors. This is done with group B-factor refinement. The ytterbium ions are each treated as a group (they will therefore undergo individual atomic B-factor refinement while the remainder of the model undergoes group B-factor refinement. The ytterbium ions are selected for refinement in the task file using:

{* select atoms in group 6 *}
{===>} group_6=(resname YB3);

      cns_solve < bgroup_yb.inp > bgroup_yb.out [2 minutes]

The B-factor refinement is followed by energy minimization of the coordinates. Only 50 steps are performed as we will follow this with a manual rebuild (see the next section). The CNS task file minimize.inp. is used:

      cns_solve < minimize.inp > minimize.out [3.5 minutes]