[gmx-users] genion deletion edges

[dsh2002 at med.cornell.edu]

Yes, genion puts it back into the box, but it deletes any atom that
protrudes out of the box. (which I guess is necessary for the periodic
boundary calculations?)
So my question now is, how far can the edge of the box be from the
molecules in the system?  I believe in principle one would like to put the
edges of the box as close as possible, but b/c my lipid bilayer has lipids
that are not exactly normal to the axes and protrude out, I have to make a
larger box that encloses the whole lipid.  So the result is empty areas on
the edges of the bilayer?

danny.


Message: 8
Date: Mon, 22 Nov 2004 11:32:36 +0100
From: David van der Spoel <spoel at xray.bmc.uu.se>
Subject: Re: [gmx-users] genion deletion edges
To: Discussion list for GROMACS users <gmx-users at gromacs.org>
Message-ID: <1101119556.18185.2.camel at vangogh>

On Mon, 2004-11-22 at 03:25 -0500, dsh2002 at med.cornell.edu wrote:
> Hello All,
> My system consists of POPC bilayer, solvent water above and below the
> membrane and a peptide in the membrane.
> I took the POPC patch from a previously equilibrated Larger patch
> originally from Tieleman, that has been simulated for some time.  So my
> patch does not have very clean edges, meaning some parts of some lipids
> protrude out of my boundaries for the box.  Will increasing the box
> dimensions so that all of the lipids totally fit in the box mess up the
> simulation by decreasing the density of the bilayer?  And does protruding
> lipids or atoms mess up the periodic boundary conditions?
> I noticed when I used genion to add Cl's to my system, the coordinate file
>  seemed to "clean" up the edges of my box by truncating ends of lipids.
> Is this a weird artifact that I have made or is this a general process of
> genion?
Doesn't genion just put it back in the box?
>
> any insights would be much appreciated.
> danny.
> _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
-- 
David.