Re: Pressure coupling options

From: Chitrak Gupta (chgupta_at_mix.wvu.edu)
Date: Sun Jun 28 2015 - 15:54:05 CDT

Hi Kevin,

This is what I usually do for my lipid simulations (note that I prepare my
lipids using charmm-gui.org, meaning that when I start my own stuff, the
lipid already has the correct area-per-lipid):

When beginning to equilibriate, I keep "useConstantArea" on. The idea is to
allow the system to adjust its density by varying only the Z-dimension (and
hold the area as it was from charmm-gui.org). Then, once the Z dimension
has stabilized (you can follow it from the *.xst file), I change to
"useConstantRatio". That way, the area is allowed to fluctuate, which is
the "normal" situation, but now, because the density is stabilized, you can
expect the area-per-lipid to fluctuate around the "correct" value.

Hope this helps,
Chitrak.

On Sat, Jun 27, 2015 at 5:36 PM, Josh Vermaas <vermaas2_at_illinois.edu> wrote:

> Hi Kevin,
>
> They do what they say on the tin. useFlexibleCell makes the unit cell
> flexible. The default arrangement for pressure control (with flexible cell
> off) changes the unit cell isotropically, useFlexibleCell decouples the
> three axes, so that the box can change dimensions independently. For
> membrane simulations, this is insufficient, as for sufficiently long
> simulations, you may get a squished membrane, as x may shrink and y may
> grow stochastically (or vice versa), and electrostatic contacts between the
> protein may reinforce this trend. useConstantRatio fixes the x and y axis
> so that they grow and shrink together, so your membrane protein doesn't
> contact itself. I am unfamiliar with how precisely its implemented, but I
> do know that I've gotten nonsense trajectories when I was learning and did
> not include those options for membrane protein systems that I got a very
> non-square shape for the membrane given enough time.
>
> -Josh
>
>
> On 6/26/15 11:32 PM, Kevin C Chan wrote:
>
> *Dear Users,*
>
> *I am a bit confused when to turn on
> “useFlexibleCell” and “useConstantRatio” options for simulations of systems
> which contain membrane (or even proteins on/inside the membrane). As the
> NAMD manual gives relatively brief descriptions, I would like to hear some
> experience sharing on this. I am interested in whether the option would
> affect any physical results and how NAMD actually achieve such constant
> ratio or area (of a membrane I guess?). *
>
> *Thanks in advance,*
> *Kevin*
> *City University of Hong Kong*
>
>
>

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:21:11 CST