# Molecular dynamics frame distorted by VMD?

I am running a VMD renderer to view my GROMACS molecular dynamics output.

At one frame, the geometry looks fairly normal like this:

At another frame, it displays a distorted geometry, with the receptor stretched out like this:

Is this common? Is it a biophysical effect of the MD system? Or is it a bug during my MD simulation steps?

If this is a problem in my processing, how can I resolve that? Or is there any setting in VMD where I can fix that?

• No, its not a biophysical effect. It's just an visualisation issue. Dec 21, 2021 at 10:59

Its neither a biophysical effect nor a bug in your MD simulation steps. It's just an visualisation issue. You could use the gmx trjconv command to make your protein and ligand to the center of the box and fit all atoms to a reference structure.

gmx trjconv -s yourTpr.tpr -f yourTrr.trr -pbc mol -center -o yourTrr_centered.trr

Clear details of individual subcommands are shown here.

• Thanks. What's the difference between TRR and XTC file by the way? Dec 30, 2021 at 7:36
• @scamander it might be a good question on the site! Comments don't last forever. Dec 30, 2021 at 7:37
• @scamander Look at this documentation.manual.gromacs.org/documentation/5.1/user-guide/… Dec 30, 2021 at 8:22
• @scamander TRR files are large and may contain velocities, energies etc. apart from the coordinates in full precision, whereas XTC files are compressed because they store only the unit cell dimensions, and the atom coordinates in reduced precision. Jan 1 at 11:58

Don't Worry, it is not a problem, This happened because you used PBC (Periodic boundary condition) and VMD sometimes has bugs during the visualization in this case. The most important to you to see that the protein wasn't broken through all frames.

You can solve this problem by using this command pbc unwrap -sel "protein" inside the Tk console as in the picture attached. Due to the PBC, the protein may have moved out of your box, so this command just ensure that the protein is in the box (by removing the frame). This will be usefull for you if you are going to do RMSD or center of mass etc.

• Just to point out, these are not caused by bugs in VMD Dec 21, 2021 at 10:58
• @Vasista Depends on your definition of bug. VMD sometimes has information about the size of the simulation box (used for displaying periodic images and the selection keyword "pbwithin"), In such cases, it could use the nearest image convention when displaying bonds and continuous protein representations. This would complicate all the drawing routines, but it could be done. The source is available so anyone so inclined could do it. Dec 21, 2021 at 18:41
• @scamander How about pbc unwrap -sel all or if it is a gmx trajectory, you could also do gmx trjconv -s yourTpr.tpr -f yourTrr.trr -pbc mol -center -o yourTrr_centered.trr Dec 22, 2021 at 3:15
• @scamander Try to use what Vasista said. From my point of view, it is not a distortion, and it will appear in the visualization whatever you do, but the most important, it will not affect your analysis, so you should not care about it. Dec 22, 2021 at 9:59
• @Vasista Your solution works. I use gmx trjconv -s md.tpr -f md.xtc -center -pbc mol -ur compact -o md_noPBC.xtc -n index.ndx. And then use md_noPBC.xtc with vmd md.gro md_noPBC.xtc -e vmd.tcl command. If you put your comment as an answer, I will accept it. Dec 30, 2021 at 4:26