[gmx-users] How to make a cluster whole

[chris.neale at utoronto.ca]

There has been some previous discussion about how to get whole  
clusters of molecules (e.g. micelles, protein complexes) for analysis  
with tools such as g_gyrate that require unbroken molecules.

The -pbc cluster command, although nice, doesn't always converge  
(http://www.gromacs.org/pipermail/gmx-users/2004-September/012070.html) and  
the nojump command requires that the .tpr has a whole cluster  
(something that is often not the case when using rhombic dodecahedron  
PBC).

The following commands work together to acheive the desired result:
1. trjconv -f a.xtc -o b.gro -pbc cluster -e 0.9 (make sure to just  
get one frame)
2. grompp -f a.mdp -c b.gro -p a.top -o b.tpr
3. trjconv -f a.xtc -o b.xtc -s b.tpr -pbc nojump