[Chimera-users] plotting RMSD vs Residue number

Eric Pettersen pett at cgl.ucsf.edu
Thu Mar 2 15:47:56 PST 2006

On Mar 2, 2006, at 7:02 AM, Isherwood, James wrote:
> Hi,
> I was wondering if there was a way to use chimera to calculate Root  
> Mean Squared Deviations
> for individual residues or backbone atoms. In order to produce a  
> graph showing the relative
> freedom of different areas of a protein.
Hi James,
	We intend to add features to make this kind of evaluation easier in  
the future.  Right now, you can only do pairwise evaluations between  
structures.  I'm going to assume that you have the pair of structures  
of interest superpositioned by some means (e.g. using the MatchMaker  

A) the structures have identical residue composition...

method A-1)  You can write a straightforward yet tedious command  
script, such as:

	echo residue 1; rmsd #0:1 #1:1
	echo residue 2; rmsd #0:2 #1:2
	echo residue 3; rmsd #0:3 #1:3

and use the 'source' or 'read' commands to execute it, or just  
File...Open (it should have a .com or .cmd extension).  The output  
will appear in the reply log, which you can save to a file.

method A-2)  You can use a slightly less tedious Python script:

	import chimera
	from Midas import rmsd
	m1, m2 = chimera.openModels.list()
	for r1, r2 in zip(m1.residues, m2.residues):
		if r1.atoms[0].surfaceCategory != "main":
			# skip ions/solvent/ligand
		print r1.oslIdent(), rmsd(r1.atoms, r2.atoms)

(in future Chimera releases, the slightly clunky "print r1.oslIdent 
()..." will just be "print r1,...")
The file should end in .py and can be run by using the command 'open'  
or the menu File...Open.  Again, the output will appear in the reply  
log and can be saved from there.

B) the structures are similar but don't have identical residue  

method B-1)  You would use the Match->Align tool (with a generous  
distance cutoff) to get as many residues aligned in the resulting  
sequence alignment as possible.  You could actually have more than  
two structures in this case, though the RMSD values you will  
eventually generate will still be pairwise with respect to a chosen  
reference structure.  Use the sequence-alignment's Structure->Assess  
Match menu item to generate a residue attribute named matchDist for  
the structures other than the one you choose to be the reference  
structure.  'matchDist' is the distance between the carbon alphas of  
corresponding residues of the reference and matched structures.  You  
can then select on or color the structures based on this attribute.   
You can also use the CalcAttr tool to save the matchDist attribute to  
a file (just have CalcAttr copy the attribute to another name as its  

method B-2)  There is no method B-2.


                         Eric Pettersen
                         UCSF Computer Graphics Lab
                         pett at cgl.ucsf.edu

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-users/attachments/20060302/4b9be0a8/attachment.html>

More information about the Chimera-users mailing list