[Chimera-users] molmap and origin index

[Tom Goddard]

Hi Catalin,

  If the voxel size, origin and map size are the same for Matlab and molmap maps, and the maps look the same only shifted relative to one another then it seems that the Matlab is simply computing the map as if the grid is shifted relative to the atoms.  It seems like it is really not handling the padding the same as Chimera molmap.

  You might try an experiment where you make a PDB file with a single atom positioned at x,y,z = 0,0,0.  Then produce the Matlab map and see where it places the single bump in the map.  To see where it lies in the grid you could show an outline box around the map in Chimera (command volume #0 showOutline true,  or use volume dialog Data Display Options pane, "Show outline box").  Does it place the atom in the exact center of the box?

	Tom


On Jul 15, 2013, at 11:12 AM, Catalin Buiu  wrote:

> Hi Tom,
> 
> yes, the Matlab script produces the same voxel size, origin and map size. The same amount of padding is added, too. The density values are different, though.
> So, the two maps have the same origin index in the Chimera volume dialog Coordinates pane, they look identical but they are still shifted away.
> 
> Thanks, Catalin
> 
> 
>  
> 
> 
> From: Tom Goddard 
> To: Catalin Buiu 
> Cc: "chimera-users at cgl.ucsf.edu" <chimera-users at cgl.ucsf.edu> 
> Sent: Monday, July 15, 2013 8:56 PM
> Subject: Re: [Chimera-users] molmap and origin index
> 
> Hi Catalin,
> 
>   The Matlab calculation of the map is where the problem lies.  I don't know what Matlab script you are using so it is hard for me to advise.  I would not expect the Matlab script to produce the same voxel size, origin and map size as the Chimera molmap command.  Do you tell Matlab the voxel size, origin and map size to calculate?  If not, why would you expect them to be the same values as the Chimera molmap command?  The molmap command adds padding around the molecule when computing the map and I wouldn't expect the Matlab script to add the same amount of padding.  So the Matlab output wouldn't have the same origin grid index.
> 
>   What you need to know is for each Matlab map grid index i,j,k what is the corresponding position x,y,z in the PDB atom coordinates used to generate the map.  Then the matlab script needs to write the MRC xorigin, yorigin, zorigin as the atom x,y,z position for grid index 0,0,0.  Alternatively if you cannot change the Matlab script, you can change the "Origin index" in the Chimera volume dialog Coordinates pane.  This value should be the i,j,k grid index that corresponds to atom coordinates x,y,z = (0,0,0) and those grid indices can be non-integer.
> 
> 	Tom
> 
> 
> 
> On Jul 14, 2013, at 7:26 AM, Catalin Buiu  wrote:
> 
>> Hi Tom,
>> 
>> thanks for your reply.
>> 
>> Now the both maps (molmap and Matlab) have the same origin index, the same voxel size and still the Matlab map is shifted away from the other map (and atomic structure). Can you tell me a possible reason for that?
>> 
>> Thanks again,
>> 
>> Catalin
>> 
>> 
>> 
>> From: Tom Goddard <goddard at sonic.net>
>> To: Catalin Buiu <cbuiu at yahoo.com> 
>> Cc: "chimera-users at cgl.ucsf.edu" <chimera-users at cgl.ucsf.edu> 
>> Sent: Monday, July 8, 2013 11:46 PM
>> Subject: Re: [Chimera-users] molmap and origin index
>> 
>> Hi Catalin,
>> 
>>   The Chimera molmap command creates a density map for a molecule using a grid that has its 3 axes aligned with the x, y, and z axes of the molecular data (atom coordinates) with padding 3 times the requested map resolution.  This is described in the molmap documentation
>> 
>> 	http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/molmap.html
>> 
>> The origin of the grid equals the x,y,z coordinates of the grid point at grid index 0,0,0.  That origin position (floating point values) is saved in an MRC 2000 file in the xorigin, yorigin, zorigin header fields. 
>> 
>> 	http://www2.mrc-lmb.cam.ac.uk/image2000.html
>> 
>> It cannot be saved in the nxstart,nystart,nzstart header fields since those are integer values and can't accomodate the exact floating point origin.  Here's the Chimera Python source code that reads MRC files if you want more details of how  that is handled.
>> 
>> 	http://plato.cgl.ucsf.edu/trac/chimera/browser/trunk/libs/VolumeData/mrc/mrc_format.py
>> 
>>   Tom
>> 
>> 
>> 
>> On Jul 8, 2013, at 1:08 PM, Catalin Buiu wrote:
>> 
>>> Hello all,
>>> 
>>> I have the following problem: when I use molmap to generate 
>>> a map from an atomic structure, the result is fine: both align.
>>> When I use another (Matlab) program to generate the map, the two,
>>> when opened in Chimera do not align anymore.
>>> My question is how molmap generates an origin index such that
>>> both the map and the structure do align and how Chimera reads (and converts)
>>> the xyz origin data from the MRC file?
>>> 
>>> Thank you.
>>> 
>>> Catalin Buiu
>>> University of Bucharest, Romania
>>> _______________________________________________
>>> Chimera-users mailing list
>>> Chimera-users at cgl.ucsf.edu
>>> http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
>> 
>> 
>> 
>> _______________________________________________
>> Chimera-users mailing list
>> Chimera-users at cgl.ucsf.edu
>> http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
> 
> 
> 
> _______________________________________________
> Chimera-users mailing list
> Chimera-users at cgl.ucsf.edu
> http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-users/attachments/20130715/7f750f1c/attachment.html>