ADL: Interesting Finding

Avni P Nimani animani at
Thu Apr 9 16:27:07 PDT 2009

Does anybody know how to account for the additional internal energy 
that occurs when flexible residues are incorporated in the binding 
pocket of the receptor?  The AutoDock4/ADT tutorial uses only two 
flexible residues, perhaps to help limit the internal energy term.   
Also, perhaps using two flexible residues of primary importance to 
ligand docking gives the most accurate Ki for docking models.  A paper 
recently published in 2008 used four flexible residues in the receptor 
binding pocket and reported the free energy change as is.  The 
contribution of the internal energy from these receptor residues can be 
quite significant.  Currently, entirely rigid model receptors seem to 
underestimate the inhibitor constant, while using five flexible 
residues seems to overestimate the inhibitor constant.  I took an 
average of the free energy changes, this should be close to the values 
of Ki found when using two to three flexible residues.  The use of two 
to three flexible residues seems to give the best approximation of an 
affinity constant so far.  Does anyone agree? 

More information about the autodock mailing list