[Dock-fans] Carbohydrate docking in DOCK 6.2

Mark J. Agostino Mark.Agostino at vcp.monash.edu.au
Wed Jul 16 19:14:22 PDT 2008

I have been trying to dock carbohydrates (2 to 4 saccharides in length) using DOCK 6.2 and am getting ridiculous results. I am seeing major steric clash between the saccharide rings. I have tried a multitude of things to fix the problem, including:
- Increasing number of minimisation steps (simplex_max_iterations, simplex_final_min_max_iterations and simplex_anchor_max_iterations)
- Turning on minimize_flexible_growth
- Changing anchor size (used 6 and 40)
- Selected more spheres for binding site (slightly expanded from cavity)
- Changed repulsive exponent in grid generation and docking run from 12 to 9 (repulsive_exponent in GRID input file and internal_energy_rep_exp in DOCK input file)
- Adjusted flex.defn to allow minimisation of some additional torsions (conj-oxygen and sp2-aromatic)
- Adjusted flex.defn to explicitly specify minimisation of glycosidic linkages (shown below)
          name         carb1
          drive_id      8
          minimize    1
          definition    C.3 ( H ) ( 2 O )
          definition    O ( C.3 ( H ) )
           name        carb2
           drive_id     8
           minimize   1
           definition   C.3 ( H ) ( 2 C. )
           definition   O ( C.3 ( H ) )
but none of this seems to generate good structures. The structures at the end look as if they have been crumpled up into a ball and placed in the binding site. I know that DOCK 4.0 used to work fine to dock carbohydrate structures, but now this seems somewhat broken in DOCK 6.2. Is there something I am overlooking here or is the problem something deeper in the program?
Thanks in advance,
Mark Agostino
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://blur.compbio.ucsf.edu/pipermail/dock-fans/attachments/20080717/d3b4adfb/attachment-0001.html 

More information about the Dock-fans mailing list