[Dock-fans] (no subject)

Terry Lang terry at cgl.ucsf.edu
Fri Sep 13 12:00:37 PDT 2002

Dear Gareth,

   First of all, I am not really clear what you mean by "take 5 spheres per 
probe with this box."  Typically, we would recommend generating spheres 
first and then building the box to encompass the spheres.  Second, are you 
using DOCK 4 or DOCK 5?  If you are using DOCK 5, make sure you model your 
input file after the DOCK 5.2.0 manual instead of the DOCK 4 manual.  The 
ligand being so far away and the short run time are both indicative of a 
problem, but I cannot be more specific without more information about what 
you are doing to prep the files.


At 07:45 AM 5/6/2005, Gareth Williams wrote:
>I'm trying to run dock on a validation set of protein:compound crystals.
>Starting with just one complex I take the protein pdb file and generate a 
>mol2 file with the openeye molcharge application (an empty line before the 
>atom records has to be deleted). I take the compound pdb file and do the 
>same (a SUBSTRUCUTE entry has to be added here). The box pdb file is such 
>that it encompasses the bound compound. I take 5 spheres per probe within 
>this box to generate a sph file.
>I set up the grid.in and dock.in files as for the first run in the DOCK4.0 
>user manual and run GRID and DOCK. The DOCK calculation takes ~0.02 
>seconds and the nrg file has the ligand 77 angstoms away from the protein 
>Any suggestions as to why the ligand is outside the box?
>Gareth Williams
>Wolfson CARD
>The Wolfson Wing
>Hodgkin Building
>Kings College London
>London SE1 1UL
>Tel:  020 7848 6806
>Fax: 020 7848 6816
>Dock-fans mailing list
>Dock-fans at docking.org

***NOTE:  This is a new e-mail address (terry at cgl.ucsf.edu)  Please update 
your address book!***

P. Therese (Downing) Lang
Kuntz and James Labs
UCSF - CCB Graduate Program
Phone:  (415)476-3986

More information about the Dock-fans mailing list