How To Perform Peptide-Protein Docking

Question

[Free Man]

Hi, could anybody give me an example about how to perform peptide-protein docking: the software and steps?
Here is my problem:
(1) a peptide about 10 amino acids
(2) target protein (an enzyme with known PDB structure file, and known active sites)
(3) how to find the most reliable binding site and the complex structure? (also need result like free energy)
Thank you!!!

Answer 1

[João Rodrigues]

You didn't really search Pubmed right?

Steps:

1. Get a structure of your peptide (make it in PyMOL)
2. Get a structure of your protein (xtal, NMR, model)
3. Use a docking program like the above.

[Fatemahakram56]

Can you please guide me that which tool is being used to make peptides

Answer 2

[Woa]

Not all Docking Software are good for protein-peptide docking. You can try ICM-DOCK(commercial) or AutoDOCK for that purpose. Rosetta might contain a module too.

[João Rodrigues]

AutoDock is a protein-ligand software, I would bet on more specific protein-peptide or protein-protein interactions. You also have FlexPepDock as an alternative.

[roy.granit]

I use PatchDock to preform such docking.

Answer 3

[Petr Ponomarenko]

I use ICM-Pro from Molsoft.com for such studies.

1) read a lot about protein family to which belongs protein of your interest and its interactions with ligand, peptides and other proteins. Is your protein a monomer, multimeter, multisubunit complex when it interacts with the peptide.

2) read a lot about domains of your protein, find which are important for this peptide binding, try to find this domains PDB structures maybe in other proteins (usually you first study domains, actually, not proteins).

3) study 3D structure properties like potential binding sites.

4) know well what you are modeling and what for. This will define your statistic function. Decide on acceptable sensitivity and specificity of your table (or you might want to optimize for AUC or something else). Know well what you are using for such modeling. Most of the software uses free energy prediction optimized for a particular task like protein folding or ligand binding, etc. Some even have multiple "free energies" for different tasks.

Answer 4

[bfmosquera1]

Hello. I am going to work with mutant receptors. So how I can determinate if the mutant receptor will cease to binding to natural ligand? What variables should I analyze?

[Petr Ponomarenko]

It is a long story =) In short, you can measure Gibbs free energy, but from good science perspective you need to study the pocket and how it interacts with different ligands and your ligand in particular.