Hi all, I have a problem with my simulations which I have performed through gromacs. Same error appeared in all simulations . VMD figure of final complex after simulation is attached. I think it is Periodic boundary conditions problem with my simulations that why water molecules are going outside during running the trajectories . I am continuously trying to solve this problem but could not success still. Is there any problem in my final md.mdp file in which i am giving for MD run? Also I have tried to solve this through trjconv option but failed . So kindly give your valuable suggestions regarding this critical situation. Specially response from Justin Lemkul,,,,,.
There is nothing wrong here at all. Molecules are "broken" in the trajectory because the physics do not depend on our visualization convenience. This is a phenomenon that has existed for many years. It is easily "fixed" by using trjconv.
http://www.gromacs.org/Documentation/Terminology/Periodic_Boundary_Conditions
I got the same problem, and I have the impression it comes from the newer gromacs version ; probably a wrong interaction with VMD.
There is nothing wrong here at all. Molecules are "broken" in the trajectory because the physics do not depend on our visualization convenience. This is a phenomenon that has existed for many years. It is easily "fixed" by using trjconv.
http://www.gromacs.org/Documentation/Terminology/Periodic_Boundary_Conditions
Dear Justin Lemkul,
Here is the figure of my protein which look likes broken protein we are confused its visual artefact produce by periodic boundary condition or really a protein was broken. kindly clear our confusion and give suggestions. How we can improve our results.
Thanks in advance.
Abid: that's the same problem as previously, a part of the protein has moved out of the PBC box and appeared on the other side of it. If you visualize it together with the PBC box it will become clear that everything is OK. trjconv from Gromacs should easily deal with it for visualization purposes.
As Justin already clearly explained, there is nothing wrong with either the simulation or VMD.
VMD has an option to see "images", "Graphics > Representations > Periodic". Select "+x" to see the neighbouring 'unit cell' along the postive x axis. (and so on)
VMD also has PBCTools extension which is very useful in 'fixing' this particular problem.
Good luck.
Just think of it like this: VMD mirrors the protein on the other side of the periodic box if it goes beyond the box.
This is a very trivial problem. Please try this method:
1. Take approximate center atom of your complex and make an entry in an index file.
then use trjconv -f md.xtc -s md.tpr -n index -pbc atom -ur compact -center -o cntr_1.xtc
It will ask 2 groups. give the first group as a center atom and output as system/complex whatever you want
2. See whether problem persists, if yes, then take cntr_1.xtc and run command below:
trjconv -f cntr_1.xtc -s md.tpr -n index -pbc mol -ur compact -center -o cntr_2.xtc
this time, use complex as a first group and system/complex as output.
(Note: The trick is you have to center it 2-3 times using appropriate atom / residue / molecule.)
- Badges
- Science topic