The first part of the molecular replacement process is the rotation search, which attempts to determine the correct rotational orientation of the search model in the cell of the unknown structure.
cns_solve < cross_rotation.inp > cross_rotation.out [26 minutes]
The fast direct rotation search is used for this rotation search. The output is a listing of rotation function peak height (in this case the correlation coefficient between squared, normalized structure factors) versus rotation angle (in Euler angles). In the ideal case there is a clear separation between the correct and incorrect solutions in the list, the correct solutions having the highest correlation coefficients. However, if the search model has low homology with the unknown structure, or the search model represents only a fraction of the asymmetric unit contents, it is not uncommon for no clear "top" solution to be present:
! mean of rotation function: 0.0218
! standard deviation around mean: 0.0153
! index, theta1, theta2, theta3, RF-function (EPSIlon= 0.25)
1 64.427 78.712 76.969 0.0799
5 64.452 56.212 47.265 0.0631
7 71.036 63.737 31.274 0.0554
8 223.333 90.000 210.105 0.0525
9 64.567 48.686 83.381 0.0509
10 54.000 78.788 324.814 0.0505
11 42.219 67.500 45.981 0.0501
13 164.255 -3.763 183.069 0.0494
15 54.896 75.025 313.167 0.0490
16 53.619 82.551 47.347 0.0478
Although the correlation coefficients are small, there is a clear gap between the top peak and the rest. This is usually a good indication that a correct solution has been obtained. In this example there is a dimer in the asymmetric unit but our search model contains only one monomer, this leads us to expect two outstanding rotation solution peaks (one for each monomer in the dimer). However, here we see only one clear peak, probably due to differences (coordinates and B-factors) in the two molecules in the asymmetric unit. The translation search will be used to determine which of the rotation solutions correspond to the two monomers.