Hi Rajiv,

You can estimate the binding free energy of docking poses by MM/PB(GBSA) method:

Article The MM/PBSA and MM/GBSA methods to estimate ligand-binding affinities

http://ambermd.org/tutorials/advanced/tutorial3/index.htm

Experimental data, showing the active site amino acids, can validate your docking poses; hence, you should check the residues, contributing to the interactions, with the experiments, then you can fairly make claim that your predicted docking poses are true or not.

1. Intuitively.

2. Decide using all available information about the protein (or its homologs), especially the effect of mutations on the protein structure/function.

3. Calculate free energies using Linear Response Approximation.

4. Determine the experimental structure of the protein-ligand complex.

Simply calculate binding free energy of the complexes, a better binding energy will tell you the correct pose. Again, before docking you may have to follow similar docking protocol (such as Grid generation, minimization, force field, H bond addition etc) to get accurate result. And also check the pdb files, whether these proteins have the similar amino acids in the active site or not.

All those free energy prediction techniques are very inaccurate. So, you will not know the right answer until the structure of the ligand-protein complex would be determined by x-rays or NMR.

I agree with Dmitri Kireev, for accurate prediction of binding pose, you need to have X-ray crystallised structure or NMR solved but, if you are using any molecular modeling techniques, then you will get probable binding pose.

Hi Rajiv,

Did you check the interaction patterns of the co-crystallised ligands with the amino acids in the respective protein chains? Are they same or different?

Hai,

As per Shabir Ahmad Ganai you can proceed. Induced Fit Docking and Metadynamics Simulations could help you to find the best pose.

Just run both the complex and find the stable conformation or pose of the ligand with reference to MM-GBSA/PBSA calculation.

Thank you all for the responses. The molecule is a kinase inhibitor and the key interactions with the hinge are retained in both the poses. But there is part of the molecule which takes a completely different pose in the two different crystal structures. This is due to a slightly different conformation which one of them adopted based on the co-crystallized ligand. I'll have to perform MD as some of you suggested.

X-ray would be best, but I don't have that facility.

All the above answers are correct to some extent. But I have a fundamental question. What do you mean by "accurate binding pose" and "correct pose"? The actual binding pose depends not only on ligand conformations but also on protein conformation. Docking only gives you a snapshot of what is possible WHEN the protein conformation is what you have used for docking, and both ligand and protein conformations continue to remain static. But proteins are known to "breathe" and their conformations vary in vivo depending on ambient conditions (water, temperature, pH, presence of other ligands etc). So what is "accurate" or "correct" in one context may not be so in another, and the context itself is a variable in vivo! X-ray structures of high quality give you the "correct pose" for a rigid conformation of the protein, which depends on artefacts of crystallization conditions (see literature on conformation of lysozyme under different crystallographic conditions). NMR structures give you a range of possible conformations in solution phase, and these conformations also depend on NMR experimental conditions. MD simulations can give you a range of conformations not found by an NMR experiment (or maybe, can be found by a number of NMR experiments under different conditions), but depends on the conditions chosen for the simulations. Finally, even assuming you have some way of knowing the "correct" pose, in vivo, the complex is unlikely to have a rigid conformation as both ligand and protein enjoy some degrees of freedom to change their conformations and binding pose, possibly in response to changing ambient conditions. So, the very idea of the existence of a single "accurate" or "correct" pose is a fallacy! Technology to observe the conformation of ligand-protein complex in vivo is yet to be developed; till then we can only make what seem to be rational and logical guesses!

Beware MM/PBSA (emphasis mine):

"Our assessment of the MM/PBSA is less optimistic. This approach appears to provide erroneous estimates of the absolute binding energies because of its incorrect entropies and the problematic treatment of electrostatic energies. ... We do not like to give the impression that our discussion of the problems with the MM/PBSA is hypothetical in nature. We strongly suggest that those who disagree with our assessment should try to validate the MM/PBSA by calculating the electrostatic energy of a buried ionized group (the relevant energetics is given by pKa measurements)." --Arieh Warshel

Article Absolute Binding Free Energy Calculations: On the Accuracy o...

First try to know the active sites of that target protein from literature as well as bioinformatics tools. After docking you can use visualization tool like discovery studio to see if that molecule fits in the active site or any other nonspecific binding.

If that compound binds with active site in both cases then you have to compare their binding energy which is the key factor for interaction.

If you have computational resources then you can go for molecular dynamics simulation study to choose the better conformation. Keep in mind that though binding energy is the main factor of docking still MD simulation study can give you a better option as it may give you an overview of in-vivo behavior of the protein-molecule interaction.

Hi,

You must look for the highest binding affinity score which is generated by AutoDock tool that you have employed for your docking. Further, if you want to get the best binding pose for your protein-ligand interaction, you can use online tools which can predict the active site binding, such as:

i) COACH

ii) COFACTOR

iii) 3dLigand Site