The scaling parameters are the most general parameters that BUSTER adjusts to optimise the fit of the model to the X-ray data.
Parameters ending _MISS will be used only if a missing-atoms channel is used in the refinement.
A large value of K_OVER, the overall scale factor indicates that the amplitudes in the MTZ file are very far from being on an absolute scale. This could be due to very low resolution of the data or because the content of the asymmetric unit (solvent content) had been estimated wrongly.
B_OVER represents a single B-factor for the model; normally it is fixed at zero and B-factors of atoms are adjusted, but if the structure is at such low resolution that it makes sense to consider a single B-factor for the whole model then it will be refined.
The B_ij terms correspond to an overall anisotropic B for the model: these terms are constrained by space group symmetry.
If B_IMPF_FRAG has refined to a large value (greater than about 5), BUSTER has determined that it has little confidence in the positions of the atoms in the model; this will often result in a very ugly correlation plot (see BRrecipCCplot wiki page) and suggests that there is something seriously wrong with the input model.
The anisotropic ratio is defined as
B_max - B_min
--------------
B_mean
(where B_min, B_max, and B_mean are computed from the overall B-factors associated with the principal axes of the anisotropic thermal ellipsoid) and is a useful measure to detect data anisotropy. Read et al.(2011) report "13% of deposited X-ray datasets have an anisotropic ratio >0.5, and only 1% have an anisotropic ratio >1".
Our STARANISO server can be used for more detailed analysis of anisotropy in your data.
Page updated Apr 2016. Please address problems, corrections and clarifications to buster-develop@globalphasing.com