[Chimera-users] about model coordinates
goddard at sonic.net
Fri Feb 20 16:48:31 PST 2015
As you say Chimera aligns your map and model correctly. So the problem is with Coot and so the Chimera list may not have an answer for you. You didn’t say the format of your map file but I guess it is MRC. In an MRC EM file the origin is typically in header fields called xorigin, yorigin, zorigin which are specific to EM data (MRC 2000 format) and are not part of the nearly identical CCP4 format used for x-ray maps that Coot is expecting. Here is some documentation on the MRC format (with floating point origin at words 50-52)
and here is some CCP4 map format documentation
For CCP4 format (used for x-ray maps) the only origin fields are called NCSTART, NRSTART, NSSTART and are integers. They are used if an x-ray map contains just a subgrid of the unit cell. The EM origin in MRC maps is floating point allowing the origin to lie between grid points. When Chimera writes an MRC map it always sets the x-ray origin to grid point 0,0,0 since Chimera map support is intended for EM maps where the xorigin,yorigin,zorigin fields are used.
So I think the following recipe is what you need to write out a new PDB file that will align in Coot. First find out the nrstart, ncstart, nsstart header vaules by opening the map in Chimera 1.10 or later and using the command “volume #0 dump true” to show the header values. If the values are -50,-50,-50 then set the origin index in Chimera to 50,50,50 (change the sign of header values): volume #0 originIndex 50,50,50 (or use the volume dialog Features / Coordinates). Now the map is positioned where Coot expects it. Now move the PDB by the same amount. Compute the difference between the new origin index and the old, say the old origin Chimera reported was 30,30,30. Then you shifted the origin by 20,20,20 grid points, in other words the map moved by -20,-20,-20 grid points. If the grid spacing is 1.2 Angstroms then you need to shift the PDB by -24,-24,-24 so use Chimera command “move -24,-24,-24 model #1 coord #1”. The save the PDB relative to the map coordinates with File / Save PDB….
This is a big pain caused by using a file format MRC that Coot does not understand. It would be much easier if you changed the map origin to say 0,0,0 so both sets of origin header fields (CCP4/MRC) agree then use MDFF with this map that MDFF and Coot will position in the same way.
> On Feb 20, 2015, at 2:54 PM, Tat Cheung Cheng wrote:
> I recently got a ~4.4A EM map and built a improved model using MDFF. The map and model is well aligned when loaded in Chimera. But it is off when I hit "center" in the volume viewer, as the origin is shifted. As I try to improve the model in Coot, the map and model isn't aligned at all in Coot, seemingly by the same offset when "center" button was hit.
> And also the values in the original index in the volume viewer is different from the values in the header in the density map. So I wonder what determines the origin index? and is it possible to find the shift and transform the coordinates in the model so as to align the model with the map in coot?
> Thanks a lot.
> Chimera-users mailing list
> Chimera-users at cgl.ucsf.edu
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Chimera-users