Re: [ccp4bb]: molrep NCS options

[Alexei Vagin]

Dear Claus,  I want to clarify the usage of lock-RF related and NCS related keywords. In the first case, we know relative orientations of molecules but know nothing of their relative positions. Thus, the keyword LOCK is relevant only to the RF. It pushes correct orientations toward the top of the list and/or improves their contrast. The impact of LOCK on the TF is only that (and when) the list of orientations is changed. This is your case. See example of using LOCK in the MOLREP tutorial at  http://www.ysbl.york.ac.uk/~alexei/molrep.html (To download the tutorial use the link “tutorial MR” in section “INSTALLATION”) The NCS and relevant keywords assume that the position of the NCS axis is known. This is unlikely your case. These keywords are used in fitting into EM maps because the point group and the position of axis are assumed before the EM maps (so called EM reconstructions) are generated. (MOLREP documentation and “tutorial EM” at the above location contain more details) Regards Alexei On 7 Sep 2006, at 15:25, Claus Kuhn wrote: ***  For details on how to be removed from this list visit the  *** ***          CCP4 home page http://www.ccp4.ac.uk         *** Dear all, I am trying to use NCS information that I gained from self-rotation function analysis for running MOLREP. My molecule obeys 72 point group symmetry with the 7-fold axis running along c, and the 2-folds perpendicular to the 7-fold along b, one of them being the crystallographic 2-fold (unit cell monoclinic C2, beta 97°, 7copies per AU. After running a self rotation function with MOLREP or any other program I can detect the 7fold axis (chi 52, 105, 157) and the 7 2-folds. I can detect the 7 rotations of my model when computing a rotation function with MOLREP as well. I am trying to input the self rotation peaks, define the NCS parameters (keywords NCS_ID, ANGLES, CENTRE) and run a locked cross rotation function (keywords LOCK, FILE_TSR and NSRF). Additionally I edit the rotation list to include just 'my 7 rotations' of the model before running a translation search What is the best way to use this information in running MOLREP ?? Is much of that redundant and may it be useless to use NCS_ID when not knowing the exact centre of the point group ? Thanks for any suggestions, /Claus AG Cramer Gene Center  University of Munich Germany