[gmx-users] Re: compressing a system

[Mark Abraham]

On 21/03/2011 11:29 AM, Moeed wrote:
> It is likely that you can use genconf to replicate a single hexane in 
> a small box (using editconf) to get something close to the right 
> density and size. This will have more hexane molecules in it, because 
> some will be removed when you use this box to solvate the PE later. 
> Equilibrate this box for a while.
>
>
>  Thank you Justin and Mark for suggestions.
>
>     Take your straight-chain PE and simulate it in vacuo for a bit
>     until you get a reasonable-looking non-extended conformation that
>     will fit in your intended box size. Use editconf to put your
>     desired box around that.
>
>
>  Is it not simpler if I take this PE in the right box size (editconf) 
> and use
>
> /genbox -cp / non-extended-PE in vacuo/.gro    -ci   Hexane.gro    -o 
>   -nmol 110 /
>
> so that I have exact number of solvet. (so no need to worry about 
> removal of some hexane after insertion PE when using command below:)

No, because genbox -ci does not achieve the kind of close-packing that 
will result in the correct density. It inserts molecules in random 
interstices - i.e. anywhere. In general, that will not achieve the 
close-packing consistent with condensed fluids. genconf allows you to 
get closer to this objective by choosing a box to tessellate, but then 
you have to equilibrate for a while to de-order the system.

It might be possible to get the right number of insertions into your box 
if you allow too much volume and then do gentle compression, but the 
presence of voids will make the system hard to keep stable, as you've 
learned. A couple of goes at genconf will be faster than that, even if 
compressing works.

Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20110321/522debee/attachment.html>