[gmx-users] -pbc nojump

[Tsjerk Wassenaar]

Hi Sara,

Please keep discussions on the list. I'm not your private tutor.

Whether you can do your analysis depends on the analysis you want to
do. But if your aim is analyzing the formation of the micelle, you're
probably better of reversing the trajectory.

> 1- trjconv -f md.trr -o md1.xtc -n index.ndx -pbc whole -s md.tpr

This makes molecules whole, which is fine. Clustering should make
molecules whole too, though, making this step redundant.

> 2- trjconv -f md1.xtc -s md.tpr -o cluster1.gro -e 600000 -pbc cluster

Fine, you get a cluster

> 3- trjconv -f cluster1.gro -s md.tpr -dump 150 -o cluster2.gro

This does nothing special. Just because you have a reference clustered
doesn't mean the output frame will turn out clustered.

> 4- grompp -f md.mdp -c cluster2.gro -o cluster1.tpr -n index.ndx
> 5- trjconv -f cluster1.gro -o cluster1.xtc -s cluster1.tpr -pbc nojump

This screws up everything. You can only use -pbc nojump with a
reference structure that is sufficiently close to the first frame of
the trajectory. Your reference is a snapshot at t=600 ns.

> 6- trjconv -f cluster1.xtc -s cluster1.tpr -pbc mol -ur compact -center -o
> cluster3.xtc

This would probably be fine if the trajectory was okay there.

Cheers,

Tsjerk

-- 
Tsjerk A. Wassenaar, Ph.D.

post-doctoral researcher
Molecular Dynamics Group
* Groningen Institute for Biomolecular Research and Biotechnology
* Zernike Institute for Advanced Materials
University of Groningen
The Netherlands