[Chimera-users] domains as ellipsoids?

Elaine Meng meng at cgl.ucsf.edu
Mon Jul 27 15:17:56 PDT 2009

Hi Stefano,
This may not help with your specific problem, but there is a new  
command to create ellipsoids for specified sets of atoms.  To use it  
you need a daily build dated July 26, 2009 or later.

It is not documented yet, but here is usage and some information:

Usage:  measure inertia atom-spec [perChain true|false] [color  

    measure inertia :1-11.a
    measure inertia #0 perChain true
    measure inertia ligand color green

"measure" and "inertia" are literal, atom-spec is a command-line atom  
specifier, the perChain option means to calculate an ellipsoid for  
every chain in the specified set of atoms.  The ellipsoid(s) will be  
shown in the graphics window as a surface model, and information  
(described below) about the ellipsoid will be sent to the Reply Log.   
This command does not tell you the relationships between different  
ellipsoids, only information about each ellipsoid individually.   
However, you can use standard vector arithmetic procedures on the  
information in the Reply Log to get angles between the vectors or  
planes of different ellipsoids.

The measure inertia command is mass weighted.  There is no option to  
weight atoms equally.  Here are more details about the output.  The  
vectors v1, v2, and v3 are the principle axes (longest to shortest).   
The axis lengths are called a, b, c (half diameter of ellipsoid along  
axes v1, v2 and v3).  And the moments are given as sqrt(inertia/mass)  
about each axis, called r1, r2 and r3.  The meaning of r1 is that it  
is the radius away from axis v1 where you could put all the mass to  
have the same moment of inertia about v1 as the molecule. So r1, r2,  
and r3 are effective radii about the axes v1, v2, v3.

There is a perChain option that makes a separate ellipsoid for each  
chain using the specified atoms.  If no color is specified it matches  
the first atom's color.

Elaine C. Meng, Ph.D.
UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab
Department of Pharmaceutical Chemistry
University of California, San Francisco

> On Jul 24, 2009, at 7:20 AM, Stefano Ciurli wrote:
>> Hello Elaine,
>> I am writing to ask you the following: I need to define the
>> relative orientation of two domains of a protein, and I am not sure
>> about how to do it. I was thinking of approximating the domains as
>> ellipsoids and then calculate the angles between the axes of the
>> ellipsoids. I wonder if you have a better way to define the
>> orientation, or if Chimera can do things like calculate ellipsoids
>> axes and their orientation.
>> Regards
>> Stefano
> Hi Stefano,
> There is an Axes tool (under Tools... Structure Analysis).  It does
> not give ellipsoids, just the long axis, shown as a cylinder.  You
> can define axes for helices or for any set of atoms (e.g. a domain),
> and then choose any two axes for an angle calculation.  For details,  
> see
> <http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/
> structuremeas/structuremeas.html#axes>
> or just click the Help button on the dialog to see your local copy of
> this information.  Eventually we want to also allow defining
> centroids and planes, but only the axes measurements are available
> currently.
> Please send chimera questions to chimera-users at cgl.ucsf.edu rather
> than to me personally, unless private data are included.  This allows
> others to benefit, or to answer the question if I am unavailable.
> I hope this helps,
> Elaine

More information about the Chimera-users mailing list