[Chimera-users] [ViewDock] wrong bonds

Elaine Meng meng at cgl.ucsf.edu
Mon Aug 29 09:45:54 PDT 2005

Hi Aurelien,
Your question is somewhat for the DOCK developers.  Our understanding
of the format is only from relatively few examples, since (like you)
I don't think there is a formal specification on the web.  I think
they comply with the official PDB and MOL2 formats except with extra
information such as scores and molecule names in comment lines.

However, there are examples from DOCK4 in the ViewDock man page:


The last section has links to the example files
(both PDB and MOL2 ligand output from DOCK4).

These short examples just contain one molecule, but essentially multiple
copies concatenated together would comprise a ligand file.  The various
descriptors and scores are in the comment lines that start with ##.

There is also a ViewDock tutorial in the Chimera User's Guide
that includes a multi-ligand example from DOCK4 in MOL2 format:

The display of your files sounds very problematic.  Without an example,
I can only guess that they might have all the ligands with the same  
number, or with the atom names in the wrong columns.

I hope this helps,

On Aug 29, 2005, at 2:36 AM, Aurelien GROSDIDIER wrote:

> Hi all !
> I'm currently trying to use the ViewDock plugin to visualize  
> results coming
> from our in house docking program, which dumps PDB files. These  
> PDBs are them
> transformed into in a DOCK3/DOCK4 compliant format. First, I should  
> say I
> haven't seen a clear description of this format, and the script  
> gathering
> together PDB files is very empirical. The problem I see might come  
> from a
> mistake of our docking program. Urls to fock3/4 specifications are  
> thus
> welcome :)
> Problem context:
> ----------------
> The problem happens when dealing with ligands having several  
> distinct residue
> numbers or if the docking coordinates do not consist only of a  
> ligand, but
> also a set of sidechains of its receptor (also with distinct  
> residue number).
> Problem symptoms:
> -----------------
> The representation loaded using ViewDock show bonds between each  
> and every
> atoms of the ligand (and the sidechains, if any), no matters their  
> atom type
> or how far they are. This is kinda of weird, as the representation  
> is perfect
> when these PDB are loaded from individual PDB files (not using  
> ViewDock).
> I am a newbie on this mailing list, and unfortunately this problem  
> should be
> solved withing a few days or so. I would like to apologize in  
> advance in case
> this help request should go somewhere else. And in case you need a  
> clearer
> description of the problem (with figures, input files and so on),  
> feel free
> to contact me either directly or using the ML :)
> Thanks a lot for your help,
> Aurélien Grosdidier
> _______________________________________________
> Chimera-users mailing list
> Chimera-users at cgl.ucsf.edu
> http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users

More information about the Chimera-users mailing list