[Chimera-users] Measure relative orientation of molecules

Tom Goddard goddard at sonic.net
Tue Jan 10 16:08:55 PST 2012

Hi Ed,

   Yeah for chains A and B in a single PDB file use

     match :.A :.B move false show true

which will report in the reply log

Motion from original file coordinates
   Matrix rotation and translation
      0.86421485   0.50312170  -0.00112044   0.02930667
     -0.50311880   0.86421463   0.00213137   0.02433383
      0.00204064  -0.00127825   0.99999710  48.44760893
   Axis  -0.00338844  -0.00314144  -0.99998932
   Axis point  -0.30430942   0.18590569   0.00000000
   Rotation angle (degrees)  30.20700574
   Shift along axis -48.44726749

This requires a Chimera daily build


since the 1 year old production release (1.5.3) does not allow matching 
two parts of the same molecule.

   The above will only work if chains A and B have identical atoms.  If 
they differ in any way you can try specifying equivalent sets of atoms, 
for example CA only,  "match :.A at CA :.B at CA move false show true".  If 
the sequences differ then you will need to use MatchMaker and it still 
requires two separate models and you could use two copies of the same 
PDB file.

   Sorry I read your original question too quickly.


> Hi,
>   There is actually no map involved. The PDB coordinate system has the 
> axis at x=0, y=0 (along z). I am not trying to fit these, only to look 
> at the angles within this PDB file.
> Regards,
> Ed
> On 1/10/2012 6:31 PM, Tom Goddard wrote:
>>> Hi,
>>>   It is not obvious to me if there is a simple way to do this in 
>>> Chimera. Given a PDB file containing a number of chains of identical 
>>> molecules arranged into a helix, but without exact helical symmetry, 
>>> is there some way to measure the rotation between adjacent chains? I 
>>> imagine that I can use MatchMaker, and then convert the resulting 
>>> rotation matrix to the angle about the helical axis...
>>> Regards,
>>> Ed
>> Hi Ed,
>>   If PDB models #1 and #2 are fit to a filament map you can use 
>> Chimera command
>>     measure rotation #1 #2
>> which will produce output in the Reply Log (menu Favorites / Reply 
>> Log) like
>> Position of 1mwk (#2) relative to 1mwk (#1) coordinates:
>>   Matrix rotation and translation
>>      0.89907669  -0.42653054  -0.09865501 -14.80260678
>>      0.38452015   0.87709721  -0.28782762   9.59677669
>>      0.20929731   0.22084426   0.95258724 -47.13790770
>>   Axis   0.50580078  -0.30621414   0.80647286
>>   Axis point  19.41814624  18.14474471  -0.00000000
>>   Rotation angle (degrees)  30.18795070
>>   Shift along axis -48.44128198
>> This specifies the rotation in PDB #1 coordinates, not the 
>> coordinates of the map these two molecules have been fit into.  It is 
>> described as a 3x3 rotation matrix and a translation (4th column of 
>> above matrix).  And it is equivalently described as an axis, point on 
>> the axis, rotation angle and shift along the axis.  The command will 
>> also show in the graphics window the axis of rotation for this latter 
>> specification.
>> Unfortunately there is not an option to express this rotation in the 
>> map coordinate system which is probably what you are interested in.  
>> Here's a way to do that though.  Fit the first molecule in the map.  
>> Then write out that PDB file using the map coordinates using Chimera 
>> menu File / Save PDB... with the "Relative to model mymap.mrc" option 
>> enabled.  Now close that fit molecule and open the saved PDB.  It has 
>> the same coordinate system as the map.  Open a second copy of that 
>> PDB and fit it in another position in the map.  Then use the "measure 
>> rotation #1 #2" command (can be shortened to "me rot #1 #2").
>> I'll see if I can easily add an option to the measure rotation 
>> command to report the result in the map coordinate system.
>>     Tom

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-users/attachments/20120110/2b20aa6e/attachment.html>

More information about the Chimera-users mailing list