[gmx-users] gmx distance

[Ming Tang]

Dear gromacs experts,

I am stretching a triple helix along z direction by fixing one end and pull the other. The helix length changes from 35nm to 50nm during the MD simulation. However, after using
gmx distance -n index.ndx -f traj.trr -s topol.tpr -select 'com of group start plus com of group end' -oav dist.xvg
to get the distance between the two end groups,   I found that the distance varies from 2nm to 15nm during the simulation.
I am quite confused about this, because the original length of the helix is around 35nm. The configuration looks correct....
Am I wrong when using gmx distance?

Thanks in advance.