[gmx-users] minimization problem

[Mark Abraham]

Hongyan Xiao wrote:
> Hi,
>  
> I has used the periodic boundary conditions and correct em.mdp 
> file. However, the minimized result is water/decane/water, while I set 
> decane/water/decane as the initial system . I cannot understand the 
> reason. Please give me some constructive suggestion! Thanks!

You need to understand that decane/water/decane under periodic boundary 
conditions has only one interface... the two decane segments are 
contiguous. Further, there's nothing magical about where you've placed 
that interface with respect to your initial box, since there are an 
infinite number of mathematically equivalent alternatives. The physics 
of the system and the algorithms in GROMACS will feel free to ignore 
your arbitrary choice. If you want to reconstruct a particular choice 
after a simulation, you will need to play around with editconf and/or 
trjconv.

Mark