I am running a protein-ligand md simulation for 100ns. The RMSD plot shows, complex remains stable upto 75ns, but exhibits very high RMSD thereafter. Upon visualizing the trajectory, I observed the ligand is no longer binding to the protein beyond 75ns, and jumping around the simulation box.
I recentered the trajectory using the following command,
gmx trjconv -s MD.tpr -f MD.xtc -o MD_center.xtc -center -pbc mol -ur compact
After recentering, I calculated RMSD, whose plot is attached below.
I also removed the periodic boundary condition with pbc -nojump flag which gave a similar plot.
I am worried what could’ve gone wrong during the simulation?
I performed the docking in autodock vina and took the best pose with lowest binding energy, and followed every step in the tutorial link
I was under the impression that the entire system (protein + ligand) should remain stable throughout the simulation, with an RMSD around 0.3–0.4 (in nm) indicating stability.
Is this kind of behavior typical in protein-ligand simulations? Would it be advisable to rerun the simulation to check if the same pattern repeats?
Protein-ligand binding is a dynamic process (in almost all cases). You would expect the ligand to leave the binding site (after a while) and rebind again later if you simulate long enough.
The binding kinetics (on/off rates) are different from one protein-ligand complex to another. In some cases it might be difficult to simulate long enough to see the unbinding event. The binding event often takes even longer in a simulation due to the larger volume of exploration.
It is always a good idea to re-run simulations (at least three replicas, but preferably five or more), unless you have a very good reason not to. The observations from one single simulation does not say much about the probabilities or time scales of an event.
Thank you for your explanations.
Yes that makes sense. I will consider rerunning the simulation with same force field and parameters to gather more reliable insights.
