I have a homodimer of a protein that I want to dock with 2 testosterones and 2 NADs, but i want that 1 testo + 1 NAD binding to the first half/chain (“A”) of the homodimer and the other to the other to the other one (“B”).
In the past i leaned how to submit homodimers to haddock and, mine are all good and renumbered for the chain B and appearing all refering (“merged”) to the chain A.
I do know which residues formed chains A and B AND i want to bind the testosterone to a known storoid binding site, imputing the position in the “active site” and I want to do the same for NAD , in a also known NAD binding site (these are known in humans and i converted to my species to do this in-silico analysis).
My problem is, that as it is a homodimer I will need to give the total of positions referring to a “copy” in both “chains” .
Yesterday I tried the following:
1 - imputing all five molecules : A- homodimer, B- testo1, C- nad 1, D- testo 2, E - NAD2
and in the imput parameters for molecule 1 I added all sites together (all storoid binding + all NAD binding) and only allowed B to be docked to C and A, C to A and B ; D to E and A, and E to D and A.
I defined the same in the interaction matrix and selected “Surface contact restraints” and submitted using the adviced values for protein-ligand docking you provide
2 and 3 - imputed 3 molecules: either homodimer + testo 1 + testo 2 OR homodimer +NAD 1 + NAD 2
where i submitted either all steroid binding sites for the 2 “chains” OR the NAD binding sites for the 2 “chains”, and again only allowed the docking to be from homodimer to testo/nad1 and from homodimer to testo/nad 2 and did the same in the interaction matrix and selected “Surface contact restraints” and submitted using the adviced values for protein-ligand docking you provide
But this is not how i want it to happen ideally, because i want the testos and NADs to bind specifically to those known reagions, one in each chain, therefore i thought of a "patchwork approach where I would do:
4 - homodimer + testo with binding sites for chain A
5 - homodimer + testo with binding sites for chain B
6 - homodimer + NAD with binding sites for chain A
7 - homodimer + NAD with binding sites for chain B
And then “merge” the best results from 4,5,6,7 in pymol
but HADDOCK blocks in the imput parameters and insists that for molecule 2 (in this case the ligand) active sites also need to be submitted …
So how can I make this more targetted approach , would be my question, so that I can really mimic what happens in biologic reality ?
I would be deeply thankful if someone could provide me with some knowledge
and thank you again for this great tool