ADL: problems with Ki values, extremes Ki

mswingle at mswingle at
Sat Apr 4 00:52:44 PDT 2009


You shouldn't expect much agreement between experimental Ki's and autodock's 
predictions - especially when using a homology model. Still, that's a huge difference you're 
seeing so there may be other issues at work here.

What does the ligand's internal energy look like in the two cases (i.e. flexible-receptor and 
rigid-receptor dockings)? Do the docked structures look reasonable? I'm also wondering if 
your total number of atom types went over 6 when you added the flexible residues. 



----- Original Message -----
From: Miguel Quiliano Meza <qilime at>
Date: Saturday, April 4, 2009 1:07 am
Subject: ADL: problems with Ki values, extremes Ki
To: autodock at

> Dear Dockers.
> I have a doubt about Ki obtained in Autodock 4. According to 
> experimental researches, my compound shows a  Ki = 2uM, but in silico 
> shows a Ki= 4yM (y=yocto). Somebody had the same experience..??
> Here some details and questions:
> * My protein was modelling by homology (identity = 44%), the active 
> site has the same residues.
> *In silico calculations, the analogous compounds keep the proportion 
> of Ki, that means a>b>c in the same way that experimental Ki, but the 
> problem as you can see are the units... uM is very different that yM.
> * yM is the result of  a flexible docking (flexible residues 2 = 
> active site).
> * A rigid docking shows a Ki= 98uM (very far of 2uM, I think).
> It is well know that try to replicate experimental Ki is very 
> difficult in docking programs, especially with proteins obtained from 
> homology modelling.
> I would like to know if I could use the information of conformation 
> (complex ligand-protein), to propose or discuss a possible mode of 
> action of my ligand.
> thanks in advance.
> Miguel Quiliano
> ________________________________________________
> --- ADL: AutoDock List  --- ---

More information about the autodock mailing list