[gmx-users] NPT for compressing the system

Justin A. Lemkul jalemkul at vt.edu
Wed Jul 14 05:22:41 CEST 2010

Moeed wrote:
> Hello,
> Thank you for suggestions. Yes simulation crashes. I am not able see the 
> whole run (1000ps). For all pressures I tired the final simulation box 
> is 2.2*2.2*2.2 nm but simulation crashes before 1000ps. The longest 
> simulation is for 120 bar (727ps). and I think that is why I am not 
> getting the _after_md.gro file after mdrun in none of the simulations. I 
> wanted to take that convoluted structure and replicate the box..

Then you need to apply better methodology.  If your system is so unstable that 
it crashes, I wouldn't trust anything it produces.

> mdrun -s PE60-P120_md.tpr -o PE60-P120_md.tpr -c PE60-P120_after_md -v >& output.mdrun_md
> As you suggested I do the runs in stepwise manner. I need still a little 

Well, what exactly are you doing?  I made a quick suggestion, but what is it 
that you're trying now?

> smaller box size than 2*2*2 but I noticed sth during these simulations. 
> When I view the trajectory in ngmx on the menu bar I can select type of 
> box. When I select triclinic or octahedron the final confromation of the 
> chain is more convoluted and fits completely in the box while in case of 
> rectangular box except for pressure of 120 bar for all other pressures I 
> tried most parts of the chain remain outside of the small box. molecules 
> is more extended...I am wondeirng if positions of atoms are calculated 
> and stored how changing the box type in trajectory viewer can affect the 
> structure?!

OK, now I'm confused.  You stated a couple of messages ago that you were 
starting with a 30-nm cubic box.  You can mess with your system to represent it 
in different ways, but none of that probably means anything.  Just because it 
looks pleasing to you doesn't mean it's real.  I'm not familiar with ngmx, but 
you could try VMD and watch the box over the whole trajectory.  Just turn on the 
box with "pbc box" in the Tk console and watch how it changes over time.

At this point, it would be useful to see the commands you used to actually build 
the system, especially if they're different from what you told us before.  It 
sounds like you might have built the box wrong, but it's hard to decide that 
when you've otherwise just described box manipulation.



Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list