Residence time in molecular dynamics simulation is not a good measure of the strength of protein-ligand interaction.

Calculate protein-ligand binding free energies and compare them with the corresponding experimental quantity.

One does not need to simulate the actual dissociation event: it is sufficient to simulate the endpoints: (1) protein-ligand complex in solution and (2) free ligand in solution.

To improve on Martin's answer, you should think about FEP and TI (free energy perturbation and thermodynamic integration). These two methods will help you with the answer that you are looking for.

Please note that these methods will not shed any light on dissociation, but let you measure binding energies.

However once you have the energies - you can get an idea regarding the dissociation of the complexes.

Hope this answer was informative.

Inspite of all the (rather trivial) tricks invented and advertised for getting

free energies out of MD.I believe the free energy problem is not satisfactorily solved yet simply because Newton did not know anything about entropies,chemical potentials and statistical physics.Metadynamics is at its best a way to sample the conformational space of biomolecules.The FEP and TI tricks mentioned by Ashar are cumbersome and may lead to irreproducible results due to non adiabatic effects,hystereses and mathematical details used,in addition there is no

guarantee that the molecule does not change conformation during the in

silico alchemy involved.

Try to use Funnel metadynamics ( "... allows the ligand to enhance the sampling of the target binding sites and its solvated states..This method leads to an efficient characterization of the binding free-energy surface and an accurate calculation of the absolute protein–ligand binding free energy.") go to:

https://sites.google.com/site/vittoriolimongelli/downloads

FEP is the most expensive but most accurate for calculating relative binding free energies for ligands that don't differ much. But obtaining accurate absolute binding free energies using FEP is a problem because the ligand must be completely annihilated and the cavity left by the annihilated ligand is not accessible to anything until after the final perturbation step.

---

Linear response approximation (LRA; Warshel) combined with linear interaction energy (LIE; Aqvist) can be used for calculating absolute or relative binding free energies:

Article Predicting Drug-Resistant Mutations of HIV Protease

Thanks everyone!

I am not sure if molecular mechanics is equipped enough to give a definite answer for this question. Its good for studying conformational space. I too use them though :)

Is there a free web server for that?

In general, the problem with "free web server" is not so much that it attracts financially poor scientists but that it attracts scientists who are not skilled at the method the web server uses; the consequence of the latter is twofold: (1) decreased overall quality of scientific research and (2) incorrect assessment of the method (to the detriment of those who use the method correctly).

Thank you for your input. Indeed, webservers offers a lot a of advantages and attactiveness to researchers with poor background in modelling and docking simulations. And because if that they cannot identify whether the server is providing reliable results.

Nevertheless, I do believe if the server reference was publish in jornals with good reputation and we validated in Casp-like situations, they are very useful.