[Chimera-users] Bug (?) with EnsembleCluster

[Elaine Meng]

Hi Francesco,

On Jan 5, 2008, at 12:34 AM, Francesco Pietra wrote:

> I was trying to carry out a cluster analysis from three mdcrd  
> trajectories from
> Amber 9.
>
> I reported here yesterday that I was unable to input more that one  
> mdcrd (in gz
> form) to Chimera. Therefore, I combined the mdcrd files with ptraj  
> in Amber 9,
> while stripping all water molecules, all lipid, and box  
> information, leaving
> only the protein with ligand.
>
> That combined mdcrd loaded perfectly into Chimera MD Movie saving a  
> pdb file
> (there was no option "save all frames" suggested by Elaine's email;  
> I did not
> select anything else than "untranformed coordinates")

As I said in previous messages, you have to view all the frames you  
want to save (actually play through the trajectory) before you can  
save them.  Even if the dialog says "N frames" the coordinates are  
not read until they are viewed.  So if you didn't view them, there is  
no "save all frames" option, it won't save multiple structures, and  
the saved PDB will not work in the ensemble tools.
>
> Opened successfully this pdb into Chimera, Tools, MD/EnsembleAnalysis,
> EnsembleCluster, select the new file, finally click on OK:
>
> that resulted in Chimera Error. This was a first trial, though I  
> have right
> taken the opportunity to submit an alleged bug from the window  
> inviting to do

Probably your file did not contain multiple structures, as explained  
above.

> so. From a perusal of Reply Log, I was unable to grasp the nature  
> of the
> failure. In particular, lacking documentation for ksdssp, I was  
> unable to run
> it (to assign the lacking secondary structure of the protein) with  
> different
> parameters from used
>
> energy cutoff -0.5
> minimum helix length 3
> minimum strand length 3
>
> and commands as from Log.
>
> Perhaps I should also mention that lack of secondary structure is  
> typical of
> the long procedure through different programs to carry our (1)  
> model building,
> (2) docking, (3) MD.

We realize this is a common situation, and that is why Chimera is  
able to calculate secondary structure automatically.  The message  
about lack of secondary structure is not an error message or  
complaint about a problem.  It is just an "informational message"  
that Chimera is automatically running ksdssp for you to calculate the  
secondary structure assignments since they weren't in the file.

Thus, you do not need to run ksdssp yourself unless you wanted to,  
and I don't think ksdssp is involved in your problems here.  However,  
you can run ksdssp with a command or from the Model Panel.   
Documentation:
http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/ksdssp.html
If you choose Help... Search Documentation from the Chimera menu and  
search for "ksdssp" that is the top hit.

I hope this clarifies things,
Elaine