Protein-Ligand Docking

[Lavande]

Dear All,

I'm a french student and I'm in practice. I try to work on docking, on
pre-docking exactly for instance.
I would like to know in practical terms how we prepare a protein
structure. Are there many methods to execute according to the protein
family ? Why do we take these decisions ? Are there different methods
following that we want to realize a rigid or flexible docking ? Are
there some softwares more efficient than others ? If yes, why ?

Many thanks.

[Alex Fan]

I have the same questions as Lavande posted.
How to make such a decision for chemist or biologist is clear for
them, because they have such experience. I'm wondering could such
experience be reproduced by other people, such as a student without
the experience?
Do we have some general rules about protein preparation? or it depends
on different type of protein?

Thanks!

Alex Fan

[Arul Mugilan]

Dear Lavande

   If you want to work in docking you should studied about the software HEX a autodock. The above two are free downloadable softwares. Just click the manuals, read and do it.

Dr.S.Arul Mugilan

[Wen Hwa Lee]

Hi Lavande,

Your questions are quite pertinent, but I am afraid that the answer would be rather long. Why don't we work together on this?

First of all - your questions sound rather directed and focused and yet you don't know much about docking, which makes me think that these are rather questions from some course that you are probably attending to. In any case, let's have a quick look on some other questions that needs answering before:

1. There are several ways of docking. Rudimentary ones will consider only shape fitting, more sophisticated ones will try to incorporate information about charges, etc. In order to calculate charges, you'll need to know, for instance, the protonation state of all residues. But then you can't find hydrogens in most of the PDB files - just because most of the crystallgoraphic data aren't high-resolution enough (not enough data to include H in the structure refinement). So here we have the first step of optimisation that it is needed

2. What about the docking method? Is it a potential map based or an explicit (all-atoms) calculation? Dependening on the method, you need to 'prepare' your receptor differently (if by 'prepare' we understand everything prior to actual docking of the small molecule in the protein model).

These are just to show you that docking is way more complex than a simple 'here is how you do it'. So I am afraid that you'll need to define better your questions. Otherwise we will spend a long time in this discussion. Why don't you have a look here: https://www.molsoft.com/gui/docking.html - this will guide you through how the program ICM do the docking. You will need to register with them (free) to look at the manual.

all the best,
L.

[Lavande]

Hello everybody,

Thank you very much for yours answers. Wen Hwa Lee, I accept your proposition. It's nice to you to help me.
In fact, I work with Alex (the second message of the discussion) who is a PhD. candidate. We want to contextualize a scientific workflow for docking. So, we want to extract some contextual elements in order to understand better how a scientist (biologist or chemist) carries out for his study. We are interested in protein - small molecule docking. We think with Alex that it exists some general rules to prepare the protein structure. For example, for a target type as kinase, there are certainly several methods to prepare this type of protein.  I would like to know protocols and why we do these actions.
I will watch your link.

Thanks a lot.

2009/3/25 Wen Hwa Lee <wenhw...@gmail.com>

[Arul Mugilan]

Hello

I am not a PhD candidate. I am a professor in Biophysics and Bioinformatics from India. I don't know who are you. I received your mail and Just i gave the information about docking. Docking is not a game. First you have to study about the fundamantal software like Hex. Then you will go to study about target identification using bioinformatics databases,  simulation, conformational analysis and how to select the molecules for docking. Ok. Finaly use the Discovery studio software because Hex is not enough for docking methods.

S.Arul Mugilan

 

[Wen Hwa Lee]

Hi, Lavande/ Alex

You are correct in assuming that there are differences in a docking protocol depending on the type of target. However the differences are not on the level of protein preparation.

Protein preparation methodology is normally dependent on the docking protocol that you'll be using. For instance - there are quite a few charged residues (e.g. Lysines and Arginines) that are normally found on the surface of the protein = accessible to solvent. And quite frequently you'll see that the side-chain atoms of such residues are not ordered in the crystal structure. As a consequence,  the authors of that structure might decide to *not* add the coordinates for the missing atoms.

Now, if these highly mobile side-chains are in your active site, you will *need* to rebuild them before any docking protocol. This is the unavoidable, and it doesn't matter which docking protocol you'll be using. But then how to best 'guess' how the side-chain will be positioned? Have a think you two and let me know.

Imagine that your protocol is only shape-fitting. Then you'll have to find the correct conformations for those side-chains. Otherwise your solution at the end will be wrong (good phrase to describe anything in bioinformatics: Rubbish in, rubbish out).

If your protocol is more sophisticated, then it is even worse: not only your 'shape' is different, but also the positions of hydrogen donor/ acceptor atoms will be different (!). You won't be able to calculate the charge properly since this depends on the direction of your charges, etc.

And, finally, wrap it back to my first sentence here: depending on the family you'll need to worry about the scoring and ranking functions - this is dependent on the kind of target/ receptor. Have a look at this paper: J Med Chem. 2005 48:6012-22 - there you'll find some ideas about scoring/ ranking functions.

Cheers!
Lee

[manojit roy]

yes HEX is not only the sufficent software for the docking......
there is also a other software like ARGUSLAB.......
by using dis u can get a more exact result.....

[wenhw...@gmail.com]