I have lots of peptides and a protein as a complex structure. Is there any way for me to calculate the binding energy of them? Basically I want to see which peptide will have stronger binding with the protein. I don't need to dock since I already had the complexes. Thank you very much!
What is available to you for this experiment? Structures for each of the peptide-protein interactions? Are all peptides binding the same receptor?
If you are interested in evaluating binding ability of a set of peptides binding to a same receptor, and have one or more structures of peptide-protein complexes available, you could use the FlexPepBind approach for this task. This approach involves modeling the structure of each of the peptide-protein complexes using the FlexPepDock approach, and then ranking them according to a (slightly tailored) scoring function.
If you can provide more details, I might be able to help better.
I have several peptides which bind to the same receptor. And we already have the docking structures. I want to calculate the binding energy for them but don't need to change the conformation (no more docking). Thank you!
How did you create the docking structures?
I suggest you look at the paper in which we introduced FlexPepBind first, where we describe how to fine-tune FlexPepDock to use it for binding prediction. You can adopt this to your system, and I will be happy to help.
London, N., Lamphear, C. L., Hougland, J. L., Fierke, C. A., & Schueler-Furman, O. (2011). Identification of a novel class of farnesylation targets by structure-based modeling of binding specificity. PLoS Computational Biology, 7(10), e1002170. doi:10.1371/journal.pcbi.1002170
one of them has x-ray structure. And rest of them just single mutation to the wild-type. So we directly use the structure. And we want to score the binding energy of them to see what is the differences. Thank you!
which protocol did you use to mutate the peptide?
The peptide which is bound in the crystal structure has an optimal backbone and side-chain conformation for that specific peptide. This might not be the case for rest of the peptides. So I will suggest that you mutate the peptide and then run FlexPepDock to refine it ( also for wt ). You can use full optimization ( -pep_refine ) or simple minimization ( -flexPepDockingMinimizeOnly ). Then you can use interface score or total score or reweighted score to rank the peptides. Just simple mutating and scoring won't be much helpful from my point of view.
I see. But why I still need to run FelxPepDock for the wild-type? I though it doesn't need any docking since it has the x-ray structure. Thank you!
I have run flexpep_prepack to optimizes the side-chains of wild-type complex. Do I still need to full optimization (-pep_refine) the wild-type complex? Thank you very much!
I am doing it like this:
prepack the structure and then full optimization, and check the I_score and total score.
Is this the right way to calculate the binding energy? Thanks!
The prepack algorithm consists of three steps:
.Move the docking partners out of contact
.Side chain optimization of each partner
.Bring the partners back to the original orientation
Running prepack protocol ensures that the side chains outside of the docking interface have a low energy conformation which is essential for scoring the decoys. So prepacking step is just for preparing the complex for further optimization unless in cases where side-chains have been previously optimized by Rosetta using the same scoring function, this step can be skipped.
So my suggestion is to prepack the wt and then use that prepacked structure for further mutation and refinement or minimization. Using different prepacked structures will cause the different mutated complexes to have different total scores and that might not be helpful for comparison of total score or reweighted score.
Minimization or refinement of the wt is also needed as it might remove some clashes present in the crystal structure and gives better score. Then the comparison of mutated and wt complex will be much more meaningful as both will be optimized using rosetta scoring function.
Thank you nawsad!