Hello everyone, I'm evaluating docking poses using MD with GROMACS, for that I have two ligands whose poses were correctly predicted by the docking (i.e. RMSD between pose and crystalographic structure < 2) and a couple of ligands with no crystal structure that bound to the same site as the former.
When I started evaluating the RMSD of the ligands in the trajectories, I noticed that even those with large movements had low RMSD. Then I learned that I was (unknowingly) fitting the calculations so translational and rotational movements were not taken into account.
After that I've been trying several combinations of commands in order to fit the trajectories to the protein backbone prior rmsd calculations. Most of the times I always got RMSD values > 9 nm.
Finally I came with:
1) gmx trjconv -s topol.tpr -f md.xtc -pbc mol -ur compact -o md_noPBC.xtc
2)gmx trjconv -s topol.tpr -f md_noPBC.xtc -center -fit rot+trans -o md_fit.xtc
(selecting protein as center and backbone to do the fit)
and this gave me rms values that started low and ended very high when calculated for the ligand and not fitting
gmx rms -s topol.tpr -f md_fit.xtc -fit none (selecting the option 13 for the ligand)
The thing is the RMSD values always start on values > 0.1 nm, is this normal? I'd expect the RMSD is 0 nm for the first frame so I'm not sure if I'm doing the right thing.
Also, even for the docking poses known to be near to the experimental structure, the RMSD escalates over 0.3 nm at 150 ps when I'd expect them to keep around the docked conformation.
Am I doing something obviously wrong in here?
Hi Edward,
You can try with VMD as GUI.
Once you have loaded your trayectory, go to Extensions > RMSD Visualizer Tool.
You should align your system first, e.g. using protein backbone as reference. Then you can calculate the rmsd of the Ligand residue (noH).
Let me know if you have problems.
Thanks Carlos, I did as you said, and it worked.
I still got a doubt, it shows 102 frames instead of the 100 frames supposed to be in the trajectory.
Where those two extra frames come from?
I assumed one of them is included as reference when I make the trajectory conversion (frame 1), and the other one I have no idea, but the ligand isn't even in the pocket is supposed to be (frame 0).
Even though, when I take the frame 2 as reference the calculations have more sense. The crystalized ligand is effectively moving outside the pocket, so I'll check that again, now knowing how to properly analyze the RMSD.
- Badges
- Science topic