GROMACS version: 2020.2
GROMACS modification: No
Here post your question
Hello,
Has anyone encountered simulations where the RMSD of the protein backbone is between 3.5 and 4.0 nm?
I’ve recently performed 3 150-ns MD simulations on 3 separate dimers of the enolase superfamily; two were the same protein but one had two mutated residues, and the other was a homologous dimer. After analyzing the RMSD of the backbone for each dimer, I found that the homologous dimer had an RMSD around 0.2 nm, while the other two had RMSD that was between 3.5 and 4.0 nm. I have attached a picture showing the graphs of all three simulations at the end of this post.
Does this mean that the simulations with RMSD values closer to 4.0 nm unstable? Why would it be so much higher compared to the RMSD for the homologous protein with a very similar sequence and structure?
What was the reference structure for the RMSD calculation? Your first frames have extremely high RMSD values, so it doesn’t appear as if the protein became unstable, RMSD was initially at a very high value.
High RMSD alone is not necessarily reason to believe the structure is unstable, but simply viewing the trajectory will show you that. Some proteins have long loop regions that are very flexible and deviate from the reference structure quickly. The RMSD skyrockets, but it doesn’t mean anything is wrong. But your case is weird because your first frames have very high RMSD.
I’m not sure if I’m answering your first question correctly but the RMSD calculation was performed on the backbone of the equilibrated structure in each of the systems.
In viewing the trajectories, the systems with high initial RMSD values do not appear to be unstable and appears to behave quite similarly to the system with the lower RMSD value.
So in the first few ps of the simulation following equilibration, the structure has an RMSD > 3.5 nm? That sounds really fishy. You would see massive structural changes immediately. Double-check what you’re using as the reference for the RMSD calculation.
If you analyze the RMSD of each monomer alone, that would work, but if you’re analyzing the RMSD of the complex, no, it is not an appropriate solution.
Use trjconv to center on one monomer to bring the whole complex (more or less) to the middle of the box. It will solve your issue.
I see. However, the trajectories I have used as input when calculating RMSD thus far have been ones in which I have already centered one monomer using trjconv.
Could this mean that the PBC is actually not the cause of these unusual RMSD values? Do you think that analyzing the RMSD of one monomer alone would give a representative value of the complex?
Have you watched the trajectory for the protein in the one you are using for analysis and visually observed what it is doing? Do you see a monomer jumping from side to side of the simulation box?
Use gmx trjconv to center the reference coordinates the same way you centered the trajectory, and use those centered coordinates as the reference for gmx rms -s.
In this case, would the reference coordinates be my md_0_150.gro file? So I could center the reference coordinates with the following command and then use the new .gro file for the RMSD calculation in the -f2 option?
