ADL: Any Advice?

Avni P Nimani animani at
Thu Apr 9 16:23:16 PDT 2009

     I am a graduate student working on his thesis involving 
docking using AutoDock 4.0 and ran into an interesting circumstance.  
The docking of the ligand using a rigid body macromolecule gives a 
rather weak Ki of 10^-4 M; however, I made 5 residues flexible in the 
macromolecule binding pocket and the strength of the Ki increased to 
10^-11 M.  Much of the binding energy is being contributed by internal 
energy.  I was wondering on how to deal with this consequence, as it 
seems that much of the increase in energy is a result of the flexible 
macrolecule residues as opposed to an increase in intermolecular 
interactions with the ligand.  What would you suggest?  And is there 
any references or literature available that addresses this issue or at 
least papers in which flexible macromolecule residues are used?  I only 
found one paper in PubMed that reported using four flexible 
macromolecule residues, but I don't think they reported how they dealt 
with the internal energy issues. 
           The protein was generated by homology modeling.  The 
structure was minimized by L-BFGS method to RMS of 0.1.  The gradient 
of 0.1 was chosen simply because the desired effect was to leave the 
model as close to crystallographic data as possible, in which major 
problems such as bond distances were fixed.  The minimized model was 
assessed for peculiarities by PROCHECK, in which the model was found to 
have no steric clashes and 99% of residues were in allowed 
conformational regions.  The final receptor protein conformation was 
highly energetically favorable. 
         Flexible receptor docking without a ligand is not possible 
since nothing is actually docking.  The rigid receptor docking shows 
the internal energy due to the ligand to be small relative to the 
intramolecular energies.  This control serves the purpose of comparison 
the best because you can approximate how much contribution to the 
internal is coming from the flexible residues in the binding pocket as 
opposed to the ligand.  I appreciate advice anyone can provide 
regarding this issue knowing that the residues of interest are not 


More information about the autodock mailing list