How to predict the binding sites and possible ligands for docking studies?

Question

[nazriyakhan85]

Hello,

I would like to predict the binding sites and possible ligands of a membrane protein for docking studies. Actually, 3D structure of my target protein is not available in the PDB. So I used homology modeling to predict the 3D structure of my target. Your suggestions would be appreciated!

Answer 1

[joseph.luttrell]

Hi. Looks like you are having a pretty similar problem that I have. It's very hard to find a webserver or program that will search possible ligands and dock them to your protein all at the same time. Almost every server requires you to already know what ligand to use beforehand and where to put it.

Here's what I found. You will need to use a combination of different servers. Since you already have the predicted 3D structure of your receptor protein but don't know the ligand or where it may bind to you can use a server like this server or some other binding site prediction program first to get a general idea of where the ligands may bind. http://www.bioinsilico.org/cgi-bin/SUPER_VORFFI/htmlVORFFI/help

Then take the predicted binding sites and your predicted 3d structure and use this server. http://cassandra.med.uniroma1.it/pepcomposer/webserver/

It will predict possible ligands that may bind to your previously predicted binding site and rank them via binding energy calculations. I'm currently looking for more servers like pepcomposer that will let you define your own binding site and search for possible ligands.

edit: Also the COACH http://zhanglab.ccmb.med.umich.edu/COACH/ webserver streamlines this process a bit and does all of these predictions in one place. It's a better fit for you since you don't have to define a binding site. For me I needed to define the binding site myself so I couldn't use it.