Reply To: Modeling phosphate ion binding site in protein.

[Anonymous]

Hi,

Here’s what I did during my third trial:-
1. First, I docked phosphate ion with autodock , to search for potential binding site and I got some 3 residues on the surface of my protein binding to phosphate ion.
2. Next, I wanted to check the binding affinity + other residues (apart from those three tat I got during autodcok run) that can interact with phosphate ion in that region. So I took the search space coordinated from autodock run and docked the ion with rosetta ligand.
3. I generated some 5000 structures and after taking the best 5% , sorted using total score I was left with only 10 structures.
4. After visulazing the docked complez in pymol, I found that ion shows interaction with only 1 amino acids (rather it should show interaction with those three amino acids which I got after autodock run). So in this case what has to be done..

Am I going wrong some where or do I have to generate more number of structures in order to get better results..

For plotting the funnel I need two parameters interface_delta and ligand_auto_rms_no_super , but I cannot use the second coz my native structure doesnot have an already docked ion in it. So in that case what shall I do ??

Also,

I want to know what all parameters are good for analysing the result is it total_score or interface_delta ?? as after checking my score file I found that the total_score for all the structure lies somewhere between 227 and 229