 |
 |
OK, the first problem is whether you are using maprot in mode 'from' or 'to'. Depending on which you are using, the operators may be the inverse of what you expect. The second problem arises from the distinction between crystallographic and non-crystallographic maps. A crystallographic map has a lattice repeat and spacegroup symmetry. A non-crystallographic map is any map which does not. (An EM reconstruction is one example). Now, suppose you rotate a map by 5 degrees. It no longer has any repeat along the unrotated lattice directions, no does it obey any crystallographic symmetry operators. So, for a start, you can no longer expect it to fit your molecule at any general position in real space. It will only fit over some limited region of crystal space. So, for example, if you are looking one cell over from your centre of rotation, everything will look wrong. Furthermore, there will of necessity be discontinous breaks somewhere in the map. For this reason, you might want to use a mask when you run maprot. But that is not enough. You need to think very carefully about what region of the map you are interested in, and how you are going to go about that process. Unfortunately, this is currently tricky and problem specific, so I can't give you an easy recipie. (It is probably possible to do it automatically - but has not been done at this point). Kevin Ronaldo Alves Pinto Nagem wrote: > Dear CCP4 users, > > I was using MAPPROT from CCP4 package to rotate/translate an electron > density map in order to superpose maps from different space group > crystals. However, when applying the same operations (euler angles + > translation OR rotation + translation matrices) into the maps and into the > pdb files (via pdbset or lsqkab) the maps and the pdb files do not > superpose. I tried simple operations such alpha=5 beta=0 gamma=0 (euler) + > tx=ty=tz=0 but the result is the same. > > Is there something I am missing? Do I have to worry about new spacegroup > definition for the reoriented maps? > > Cheers > > Ronaldo. > >