[gmx-users] Re: Re: problems with some calculated trajectories

[Mark Abraham]

Bernhard Knapp wrote:
> 1) trjconv with the "-pbc nojump" option does not help - it removes most 

If I understand correctly, -pbc nojump will help when the relevant atoms 
are part of the same [moleculetype]. One fragment in each of your two 
examples is apparently the same four-strand sheet with helix wrapped 
around it. This suggests to me that you might have several chains, each 
as a different [moleculetype] and your observations are actually what 
you should expect. If so, -pbc cluster is the kind of thing you should 
be using.

In any case, a description of the composition of your system and copies 
of a series of commands that reproduces your result would always have 
been appropriate :-)

Mark