[gmx-users] multiple molecules simulations

[Gloria Saracino]

Dear all, 

I have performed a simulation on eight identical peptides (composed by 11 residues) embedded in explicit water. Box dimensions (cube of 7.92nm) have been chosen to get a high concentration to accelerate the aggregation process. PME has been used and rvdw=rcoulomb=0.9. 
The trajectory has been centered on a residue of a chain with
trjconv -f *.xtc -center -pbc mol -s *.tpr -n *.ndx
During the first 16ns even if each peptide cannot see its periodic image, the whole set of peptides (Protein in the index file) see itself in some frames at distances below 2nm, and in a very few of them below 0.9nm.
Looking at the trajectory in vmd after less then 2ns I see the formation of an oligomer composed by six peptides, the other two peptides move around leaving the oligomer on a side and approaching on an other side. The oligomer never sees its periodic image as well as the two peptides never see their periodic image. The pi violations observed for the whole system correspond to the two peptides that, together or one at time, are placed between the oligomer and its pi. Considering 0.9nm as the distance below of which there is a direct interaction I found that when the minimum distance of the two peptides from one side of the oligomer is close to 0.9, the minimum distance from the other side is always above 1.4nm.
Moreover the LJ and coulomb trends don't show abnormalities.
Even if in simulations of a single molecule the effect of periodic image violation is a clear signal of a too small box to approximate a condition of infinite diluition, how I have to interpret such a violation for a system composed by multiple molecules and in which I want to reproduce a condition of high concentration? 
Can I use this simulation to study the behavior of the system at the chosen concentration? 
There are particular simulation settings or checks that I have to take into account to handle a high concentrated solution a
nd that I overlooked?

Any help will be appreciated (I apologize if some questions may seem trivial),

Gloria
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20111124/352dbada/attachment.html>