[Chimera-users] Tube representation picture of the eigenvector

Elaine Meng meng at cgl.ucsf.edu
Tue May 28 10:28:33 PDT 2013

Hi Tom,
The ViewMotions webserver does basically what you describe, plus the initial superposition of the structures.  Although the sessions from that server show licorice ribbon (I believe), it might work to hide ribbon and show CA trace instead, if the user prefers a more angular appearance.

As you said, it is technically possible to generate something similar in Chimera given two input structures, with morphing and writing out the trajectory and reading it back in as multiple structures. However, neither ViewMotions nor Chimera compute eigenvectors.  That's the main reason I suspected VMD.

On May 28, 2013, at 10:17 AM, Tom Goddard <goddard at sonic.net> wrote:

> Hi Elaine,
>  A close look at that figure reveals that it is 20 stick models showing CA-traces side by side giving it the appearance of a flat sheet.  The CA positions move in a straight line from initial to final position.  Not sure if morph conformations can produce those linear interpolations.  If it could, and the morph frames were written out as separate PDBs and read back in to allow displaying them all at the same time, and rainbow model used, it could be possible to make that figure.  Needs testing to know if this works.
> 	Tom
> On May 28, 2013, at 10:04 AM, Elaine Meng <meng at cgl.ucsf.edu> wrote:
>> Hi Prasanna,
>> Although Chimera is credited, I am skeptical that Chimera was really used to make this figure.  As far as I know it's not possible to make something exactly like that in Chimera.  Maybe they really used VMD.  You could try asking the authors of the paper about it.
>> For others on the list, the figure in question is:
>> <http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2479573/figure/fig5/>
>> The most similar thing I can think of is to use the ViewMotions Rainbow web server:
>> <http://viewmotions.bc.edu/>
>> It takes your two structures and outputs a whole series of intermediate interpolated structures with rainbow color-coding as a Chimera session.  You can open the session in Chimera and adjust the view (rotate, scale, etc.) to your liking before saving an image.  This server is relatively new (2013) and was not available in 2008 when the figure discussed above was published.  It simply shows the differences between the two structures.  It would not necessarily be an eigenvector unless your method of creating the structures specifically makes them different only along an eigenvector.
>> The other possibility in Chimera is to show a tube with circular cross-section proportional to the CA-CA distance between the two structures when they are superimposed.  This is really a tube, round not flat, and not colored differently from side to side as in the figure and the ViewMotions result.  Also it is only shown on one structure ( you have to pick which one) and thus does not really show the exact location of the other structure. To make that type of figure, you would:
>> (a) open the two structures in Chimera, superimpose them with "mm #0 #1 show true" which would also give a window with sequence alignment
>> (b) in the sequence window, there will be an RMSD histogram above the sequences.  For two structures, this RMSD is the same as the CA-CA distance.
>> (c ) in main Chimera, menu, choose "Tools…. Structure Analysis… Render by Attribute" and in that dialog, choose attribute of "residues" named "mavRMSD", and use the "Worms" section.  You would just apply this to one structure and hide the other one. More details on using this tool:
>> <http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/render/render.html#render>
>> This Chimera approach could be based on an ensemble of structures (such as from NMR) and mainly shows which areas are more or less conformationally variable.
>> I hope this helps,
>> Elaine
>> On May 28, 2013, at 8:48 AM, Prasanna Venkatesh <prasannavenkatesh1 at gmail.com> wrote:
>>> Dear Chimera users and supporters,
>>>  I would like to generate a tube representation picture similar to one in the article (Figure 5 in "MD Simulations of Papillomavirus DNA-E2 Protein Complexes Hints at a Protein Structural Code for DNA Deformation"). Through Gromacs, I have average.pdb and extreme.pdb of my protein for the MD simulations. Please help me by providing the steps to make the picture. I asked the question earlier but not getting the similar picture with the answer provided. I don't know I asked it right the last time. Thank you in advance.
>>> Regards,
>>> Prasanna.

More information about the Chimera-users mailing list