[gmx-users] Re: genconf and density of system

[Thomas Schlesier]

Justin is right, the density of the real system does not always match 
the density of your force-field. If you look into the publication of the 
GROMOS 53a5 / 53a6 force-field you will see that.
For 1-butanol you get with 45a3 force-field 708kg/m^3, with 53a5 ff 
774kg/m3, and experiment tells you 806kg/m3 ...

The following command you tried
genbox –cp Hexane.gro –cs HExane.gro –p Hexane.top –o solvated.pdb
didn't work because you put your second hexane ontop of the first one.

With the option -cs genbox will use the -cs box and placeses it into the 
  -cp box. So if the -cs box is 1*1*1 nm, and you put it in a 3*3*3 box, 
you will end with 27 -cs boxes which sit next to each other, so that the 
whole box is 3*3*3 great.

Best thing you can do is, make a NPT simulation (look for the right 
compresibility) and look how much your density is off,if it's only a 
little your fine.

Greetings
Thomas


Message: 3 Date: Thu, 13 May 2010 17:07:31 -0400 From: "Justin A. 
Lemkul" <jalemkul at vt.edu> Subject: Re: [gmx-users] genconf and density 
of system To: Discussion list for GROMACS users <gmx-users at gromacs.org> 
Message-ID: <4BEC6A13.4010801 at vt.edu> Content-Type: text/plain; 
charset=ISO-8859-1; format=flowed Moeed wrote:
 > > Hello,
 > >
 > > I intend to generate a stack of molecules with the command line:
 > >
 > > genconf -f Hexane.gro -nbox 8 6 4 -o Hexane-STACK.gro
 > >
 > >
 > > Hexane.gro has box size of 3*3*3 nm3 and I need to have about 7
 > > molecules per nm3. So I multiplied 7 by 27 (box size) to get 
(7*27=)189
 > > molecuels. What I get from the above command is (-nbox 8 6 4 )192 but

Are you sure of that density?  If so, your box dimensions give 27 nm3, 
implying
that you'd have to put all 189 hexane molecules in that 3x3x3 box. 
Seems a bit
crowded to me, at least for a molecule of that size.

 > > the system size in getting bigger as well!.  * 24.00000  18.00000
 > > 12.00000 nm3.*
 > >

The only thing that genconf does is build replicates of a system, it 
does not
somehow intelligently pack your system together for you, so these 
dimensions are
exactly what you would expect: (8*3), (6*3), (4*3).

 > > Could you dear experts guide me how I can reach the system density I
 > > want. Also to I wanted to know what is best value -dist option. Is 
it ok
 > > if I leave the distance between boxes 0 0 0?
 > >

I've already told you a bit about the meaning (or lack thereof) in the 
density
of the initial configuration.  You could conceivably simulate under a 
constant
volume ensemble once you have a suitable configuration, but then I'd 
argue that
you're forcing the density to be right and you're not using appropriate
conditions.

You can leave -dist set to zero, since the molecules are already in some 
sort of
box, and the -dist option just adds additional space between them, which is
probably not necessary.

The easiest method for building your system may simply be to decide on a 
box
size, determine the number of molecules that fit in it, and run:

genbox -ci hexane.gro -nmol (whatever) -box (whatever)

If all your molecules can be placed in the box, genbox will do it for you.

-Justin