[Chimera-users] Questions about homology modelling

[Elaine Meng]

Hi Jim and Dmitri,
Since your questions are related, I will attempt a combined answer:

(1) structure editing.  I would not recommend Chimera for full- 
fledged homology modeling, but we do have excellent capabilities for  
mutating a few amino acids here and there and placing reasonable  
rotamers from backbone-dependent or backbone-independent  
libraries.    This can be done with the Rotamers tool (under Tools...  
Structure Editing):
http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/rotamers/ 
framerot.html

or its command implementation, "swapaa":
http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/swapaa.html

There is also an "addaa" command for extending peptides C-terminally:
http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/addaa.html

You will see other tools such as Build Structure and Minimize  
Structure (also implemented as the command "minimize") in the same  
menu as Rotamers.  These tools are not as fully developed as  
Rotamers, but we continue to improve them.  Each tool is documented;  
one way to see a tool's man page is to start it and click the Help  
button.  Likewise, each command is documented, and one way to see its  
man page is use the command "help [command]" - for example, "help addaa"

(2) interface to true homology modeling software.  An interface to  
Modeller is a future direction, but little has been done so far.   
Actually, there is Model Loops tool (again, under Tools... Structure  
Editing) that interfaces with a separately obtained copy of Modeller,  
but currently it only will model alternative conformations of loops  
that are already present in your structure.  What we do have now is  
the ability to fetch directly from Modbase from the Chimera command  
line, as described here:
http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/open.html

for example, command "open modbase:b2ar_human" today fetches 4  
different models of the human beta2-adrenergic receptor as models  
#0.1-0.4 in Chimera.  The modbase structures have the modeller error  
profile value in the B-factor column, so coloring by B-factor for  
these structures is really coloring by error profile value.  Which  
brings me to...

(3) coloring by attribute value.  Actually I believe the flexibility  
of this feature is a selling point of Chimera, so am somewhat  
surprised by your comment.  Do you mean the coloring itself is  
limited, or the choice of attributes that can be shown with color?

- If the coloring, I should note that you can move the sliders on the  
histogram, change their colors to whatever you want, and add more  
color sliders to define the mapping.  I guess the main limitation  
there is that the interpolation between the sliders is linear only,  
but that still leaves quite an array of possibilities.  See the  
Render by Attribute documentation:
http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/render/ 
render.html

- If the attributes available, there is a Define Attributes tool with  
which you can read in your own arbitrary attributes of atoms,  
residues, or models (using a simple text format), and an Attribute  
Calculator tool with which you can combine existing numerical  
attributes with formulae to generate new ones.
http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/defineattrib/ 
defineattrib.html
http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/calculator/ 
calculator.html

Linked to the Define Attribute page are some example files, including  
ones that define additional
hydrophobicity scales.  There is also an Attributes tutorial that  
uses several attributes-related tools:
http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/tutorials/ 
attributes.html

I hope this answers both of your questions, but feel free to write  
back if I missed something or was unclear - Best,
Elaine
-----
Elaine C. Meng, Ph.D.                          meng at cgl.ucsf.edu
UCSF Computer Graphics Lab and Babbitt Lab
Department of Pharmaceutical Chemistry
University of California, San Francisco
                      http://www.cgl.ucsf.edu/home/meng/index.html

On Jan 14, 2008, at 11:57 AM, Nicholson, Jim wrote:
> Hi:
>
> I still love Chimera (if you remember my last message), but, I need  
> to do some homology modelling.  I was using Insight/Homology, but,  
> access to that software has been taken away.  I’m wondering if  
> anyone on the Chimera project is using an interface to Modeller or  
> has any recommendations commands that might be usefull, like adding  
> amino acids from a sequence, extending the protein, etc.
>
> Also, the coloring schemes I can see look a little limited.  I  
> regularly color by charge or hydrophobicity.  I was wondering if  
> there might be something I’m missing in Chimera?
>
>  -Jim Nicholson
>
> Program in Neuroscience
>
> University of Maryland
>
> 685 W. Baltimore, St., MSTF 5-11
>
> Baltimore, MD 21201
>
> 410-706-3418
>
> 410-804-5872 (cell)
>
and Dmitri wrote:

> Dear friends and colleagues,
> I have not really tried the product yet, but was wondering whether it
> is possible to "edit" a protein's structure by introducing one or two
> single mutations with the follow-up minimization using Chimera. If so,
> could you please hint to the  algorithm of actions.
> Thanks a lot in advance! The project is burning :-(
> Dmitri