ADL: Theory in Vina tutorial video?

mswingle at mswingle at
Thu Apr 16 03:58:17 PDT 2009

----- Original Message -----
From: mcmartin at
Date: Thursday, April 16, 2009 3:33 am
Subject: ADL: Theory in Vina tutorial video?
To: autodock at

> Hi all,
> So I'm watching the vina tutorial video and I got a little puzzled 
> with the theory explanation. They say:
> "It is often said that the global minimum corresponds to the native
> conformation, however this is not correct, a deeper but narrower peak 
> may actually be less preferable, because of entropy"
> Oleg says this while pointing at the conformation of lowest energy,
> which as I have understood, is the "strongest" binding energy. There 
> are a couple of things in this sentence that I do not understand and I hope 
> you can help me out:
> 1. Why does he talk about a ligand-receptor complex with a "native"
> conformation? I thought the term native conformation was strictly 
> applied to proteins with the proper functional folding?

I think this is just intended to be a convenient shorthand for the "true" conformation that would, presumably, be observed 
in an NMR or crystal structure of the complex, if such was available. Native protein conformation is to biologically active 
protein structure as native ligand conformation is to the "active" ligand structure that's "competent" for binding.

> 2. So I understand in the in silico experiment entropy is lost upon
> ligand-receptor binding,  especially since AutoDock neglects the solvent
> molecules (Unless the entropy of the desolvation state is taken into account
> with the AMBER force fields, which is not yet clear to me, even though 
> I have read Cornell et al. 1995) So why is a _wider_ peak more 
> preferable in terms of entropy? In other words, what does the width of the peak
> represent?

Desolvation effects ARE modeled approximately in autodock's scoring function. I believe what Oleg is getting at is that, in 
addition to increasing entropy by displacing bound water, ligand binding is associated with a decrease in the 
conformational entropy of the ligand (and probably of the receptor as well in most cases). That is, the free ligand may 
readily take on many different low energy conformations whereas the range of bound conformations would be severely 
restricted by the requirement to maintain proper energetically favorable contacts with the receptor. A "narrow" peak in 
this context means one where the energy only stays low within some small region (in terms of the hypervolume defined by 
the ligand's state variables - translations, rotations, torsions) and then increases steeply with further changes in the 
ligand's state variables, thus limiting the size  of the conformational ensemble (in terms of hypervolume) that can fit near 
the energy minimum. A "wide" peak would have a larger low energy region (and possibly shallower gradients up the sides), 
which could accommodate a larger ensemble and, therefore, would be associated with a smaller decrease in 
conformational entropy. A smaller decrease in conformational entropy would mean, all other things being equal, a lower 
free energy of binding.



More information about the autodock mailing list