The Molecular Graphics Laboratory Forum

internal cavity
Page 1 of 1

Author:  fglaser [ Tue Jun 22, 2010 8:49 am ]
Post subject:  internal cavity

Hello all,

I have a protein whith a binding pocket completely buried into it (a cavity), so I wonder if what / how / if it is possible to use autodock in this case.

Thanks a lot,


Author:  stefano [ Fri Jul 16, 2010 5:49 pm ]
Post subject:  Re: internal cavity

Hi Fabian,

yes, you can use AutoDock, as far as there's enough space for your ligands to fit into the pocket. If not, you will get positive energies because AutoDock doesn't change the binding site conformation.

As an option, you could consider to make one or more key residues flexible, in order to allow a limited adaptation to your target.


p.s. I've moved the topic to the more appropriate AutoDock forum.

Author:  fglaser10 [ Tue Sep 14, 2010 5:17 pm ]
Post subject:  Re: internal cavity

Dear Stephano,

Thanks for your reply. I wonder how you would choose the flexible residues, I mean the cavity is small, but the channel leading to it is quite long, and I am unsure how many flexible residues Vina can deal with successfully... (10?). I thought about selecting all of them along the channel and making them flexible is that a good idea?

Thanks a lot!


Author:  stefano [ Wed Sep 15, 2010 5:14 pm ]
Post subject:  Re: internal cavity

Hi Fabian,

Depending on which program you're actually using (AutoDock or AutoDock Vina), the numbers are different.
With AutoDock, you want to keep the number of rotatable bonds of both ligand and flexible residues <= 32, while with Vina this limit doesn't exist. With both, though, you want to limit residues considered flexible because of the increase in complexity of your search.

I'm not sure if this advice is pertinent to your question, but with both grid box placement and flexible residues, you want to consider only the region of the binding site where your ligand actually binds.
So even for a long channel, you don't have to include the entire path for reaching the binding site, but focus only on the residues around a given distance (5-10 Angstrom?) around the ligand.

Hope this helps,


Author:  fglaser10 [ Sun Sep 19, 2010 2:45 pm ]
Post subject:  Re: internal cavity

Hi Stefano,

Thanks for your email, indeed it helps. I did one docking experiment, and the results are encouraging, the channel I mentioned is visible in blue (in vina2.png figure). The attached file (vina5.png) shows the docking results, in red the crystallized ligand, in green the docking experiment, without any flexible residues. The docking results agrees with two of the crystal structure positions as you can see, but what we are really interested in is to investigate a mutation near the second ligand (in red) starting from the bottom. For which the free docking did not find any suitable result.

So my next question is how can I "force" in some way the docking of my ligands to the position in the channel which is nearest to the mutation point (in cyan in vina7.png, the green cluster of docking results above it are not relevant, are outside the channel), or what is the best way to tackle this question at all: We have experimental values for 5 ligands and 4 mutations on the cyan spot, and I wonder what is the right way to go. Since I don't have much docking experience any suggestion will be highly appreciated.

Thanks a lot!!

Fabian Glaser
Technion, Israel

File comment: mutation location in cyan
1t0s_vina7.png [ 64.48 KiB | Viewed 13634 times ]
File comment: channel
1t0s_vina2.png [ 73.44 KiB | Viewed 13634 times ]
File comment: docking experiment
1t0s_vina5.png [ 37.34 KiB | Viewed 13634 times ]

Author:  cyndarelli [ Sat Oct 02, 2010 1:43 am ]
Post subject:  Re: internal cavity

I had a very similar question, thanks! this answered it.

Page 1 of 1 All times are UTC
Powered by phpBB® Forum Software © phpBB Group