[Chimera-users] Joining fragments in a peptide bond

Elaine Meng meng at cgl.ucsf.edu
Mon Sep 20 16:06:34 PDT 2010

Hi Sujata,
I'll answer the easy part first: you can specify the axis cylinder radius as an Angstrom value in the Define Axis dialog (opened from the Axes/Planes/Centroids tool) or if you are using the "define" command, with the "radius" keyword, see the general options in that page:

I see you have a left-handed polyproline helix which also has a superhelical twist, and you wish to extend its length maintaining the same helicity and superhelicity.  Then you have assembled three of these into a triple helix.  This problem is not really something you want to attack with Build Structure; instead, you need to figure out the symmetry operations to extend this structure, then apply them to additional copies of the identical structure to extend it.

One way would be to figure out BIOMT matrices and add those to the PDB file.  

However, it can be difficult to figure out the correct values and edit them in using the correct format, so I'll describe a different way that just uses commands.

The two fragments you sent are identical (I know you know this, I'm just explaining to the others), and I matched the first half of one copy to the second half of the other copy to try to figure out the symmetry operations.  Command:

match #1:1-15 at ca #0:16-30 at ca


This match looked pretty good to me, so next I used the measure command to tell me the transformation that had been used to generate this match:

measure rotation #0 #1


Actually that command will display another axis cylinder, and I don't think you can control the size of that one, but you can just close or hide it with the Model Panel (under Favorites in the menu).  The transformation information is given in the Reply Log (also under Favorites in the menu):

Position of #1 relative to #0 coordinates:
  Matrix rotation and translation
    -0.00777505  -0.89851915  -0.43886545  45.94850252
     0.89297961  -0.20375330   0.40133777  -0.02182236
    -0.45002996  -0.38877748   0.80394347 -26.28870812
  Axis  -0.40352597   0.00570191   0.91495042
  Axis point  17.07887792  12.85324577   0.00000000
  Rotation angle (degrees) 101.75881858
  Shift along axis -42.59440324

Then I reset so that the two fragments were again exactly superimposed (since they are identical ), using the command:



Then I applied TWICE the reported translation and rotation because the match was of 15 residues, but the whole fragment is 30 and I want to join the ends.  I used the axis given in the Reply Log as the axis of movement and the axis point as the center of rotation, and applied twice the rotation angle with the "turn" command and twice the shift with the "move" command:

turn   -0.40352597,0.00570191,0.91495042 203.5 models #1 coord 0 center 17.07887792,12.85324577,0
move  -0.40352597,0.00570191,0.91495042 85.188 models #1 coord 0


At this point it looks like they are in a pretty good position to just add the bond and delete the extra atoms at the ends.   Since you already have the models in position, don't use Join Models in Build Structure.  Instead use "copy/combine" in the Model Panel (or the "combine" command) to merge the two fragments to create another model, and in that model, select the extra atoms that you want to delete (probably all hydrogens on N-term N and one of the oxygens on the C-term carboxyl of the other fragment), use command "del sel" to delete them, then select the two atoms to be bonded, and use command "bond sel".  The bond will look a little strained, but this whole approach is going to get you an APPROXIMATE model.  You could use command "addh" if you cared about the presence of the amide hydrogen at the join.   This approximate model is good enough for a figure, or as a starting point for a simulation in some other package like Amber.


Actually, it might have been better to start with the trimeric fragment with all 3 chains, use residues from all 3 chains for the match, etc., and then there would be three bonds to form, rather than repeating only the single strand and then trimerizing.  Hard to know without trying.  Sorry, it is a rather laborious process.

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

On Sep 20, 2010, at 10:31 AM, Sujata wrote:

> Hello Elaine,
> Thank you for the email below. It was very helpful.
> But, I would like the second fragment to be on the same helical axis as the first one. When I join them this way, by defining the torsion angle, the second one does not stay on the same axis as the first one. So is there a way that I can fix the two peptides to be on the same axis when I fix the bond length and the torsion?
> I am attaching the two individual files and the combined one.
> Also when one defines axis, is there a way to reduce the thickness of the displayed axis. Due to its thickness, the atoms are not visible.
>  Thank you.
> Regards,
> Sujata

More information about the Chimera-users mailing list