I suspect that it is a visualization problem that comes from vmd or a similar program. You should remove the jumps and the rotation and translation. To do that , use the trjconv nojump and rot+trans options.
Hi Akanksha, Im not a gromacs user but it sounds like a visualization problem more than the system itself. if you use periodic boundary conditions this should be solved just centering the trajectory. You notice changes in the total energy of the system or something unusual?