[gmx-users] switching from Group to Verlet: broken atoms on the box border

[andrea]

Hi there,

I have performed the equilibration steps of a system protein-water using 
as cutoff-scheme Group. Since I need to carry out the production run on 
a GPU cluster, I needed to switch to Verlet scheme. Visualizing the 
trajectory from the production run on vmd the atoms on the box border 
are "broken" (the water molecules). This happens immediately after the 
first step of the simulation. And this happens on CPU cluster too. 
Trjconv -pbc whole fix this "problem".
I know that the best thing is to start with a scheme from the beginning 
to the end but I am wondering whether this switching it is safe.
I am thinking to save this simulation taking the output from the 
equilibration using the Group scheme, switching to Verlet scheme, 
equilibrate for few ps (i.e. 300ps) and then use the later output, after 
trjconv -pbc (in order to get a proper gro), for the production run.

gromacs version 4.6.3

thanks for any help

and

-- 
Andrea Spitaleri PhD
D3 - Drug Discovery & Development
Istituto Italiano di Tecnologia
Via Morego, 30 16163 Genova
cell: +39 3485188790