[gmx-users] compression to study pressure effect

[Justin A. Lemkul]

Elisabeth wrote:
> Dear all,
> 
> I am trying to study the effect of pressure on total potential of my 
> system (8 polymer chains). My problem is that I dont see a systematic 
> effect of pressure on potentials and I cant judge if different pressures 
> increase or decrease potential. This is the critical observable in my 
> system and with high fluctuations I am getting I cant make comment on 
> pressure effect. Total drift is high in most potential functions..I 
> start my runs from a frame which is the output of an older NPT run (I 
> use cpt file) that has a close density to what I want. In the production 
> runs (NPT for 12 27 70 bar).
> 

How large are these polymers?  If they're relatively big, their dynamics will be 
slow and high pressure may just mash them together in some physically 
unreasonable way.  Have you watched the trajectories?  What type of behavior do 
you observe?

> Below is the settings I am using and I really appreciate it if you could 
> comment on the most important factors for such a study. I tried 
> different cutoffs as well...I though maybe increasing cutoffs invloves 
> more interactions and can represent better the pressure effect...
> 

Increasing cutoffs often leads to *less* accurate and otherwise artifactual 
behavior.  The cutoff values, electrostatic scheme, etc are generally 
pre-defined for the force field you're using.  The values you have set below 
seem very arbitrary and not likely to correspond to a reasonable physical model.

<snip>

> Statistics over 2043101 steps [ 500.0000 through 4586.2000 ps ], 17 data 
> sets

<snip>

> Statistics over 750001 steps [ 500.0000 through 2000.0000 ps ], 17 data sets

<snip>

> Statistics over 2750001 steps [ 500.0000 through 6000.0000 ps ], 17 data 
> sets
> 

All of these timeframes are very short, so your results probably aren't very 
well-converged and your system may still be equilibrating.  Only for very small 
molecules (i.e., water and other small things) will you gather reasonable data 
in such short times.

-Justin

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================