[Dock-fans] Re: how to solving segmenation fault while docking

Scott Brozell sbrozell at scripps.edu
Tue Mar 13 14:10:28 PST 2007


Please send questions to the list instead of to individual developers.

On Thu, 8 Mar 2007, kureeckal ramesh wrote:

>   I am using academic version of DOCK5 (installed on to FEDORACORE 5 version) for docking few ligands onto the homology modelled protein structure. I was sucessful in docking few ligands onto this structure (with max of 100 spheres).

Upgrade to DOCK 6.1.  Many memory related problems were fixed in the latter
versions of DOCK 5 and into DOCK 6.

>   Though the file 'sph' could generate a number of cluster with different spheres, I am getting segmentation fault repeatedly, for a new ligand that I want to dock onto this protein. This error is observed while using different combinations of spheres that was selected from the '.sph' file. I even tried using 228 spheres (max number from 1st cluster), but this is also not working.

What program is seg faulting ?  What input files have changed ?
What are the differences between the old ligands and the new ligand ?
What happens when you use the old sphere representation of the active site
with the new ligand ?

These are the types of questions that you can use to divide and conquer
the problem.  By making incremental modifications starting from a
working calculation, the problem can be localized.

>   I have also seen the sphgen_cluster.pdb file that was generated along with the protein structure using chimera package.  Can this be of any use ?

Definitely, you should scrutinize everything that has changed;
visualization is one of the quickest ways to see something obvious.

>   I request you to kindy address this problem for which I shall be thankful.
>   Yours sincerly
>   Ramesh K V

More information about the Dock-fans mailing list