AMBER Archive (2008)

Subject: RE: AMBER: Protein rotating out of box

From: Steve Spronk (spronk_at_umich.edu)
Date: Wed Apr 02 2008 - 08:13:12 CDT

> -----Original Message-----
> From: owner-amber_at_scripps.edu [mailto:owner-amber_at_scripps.edu] On Behalf
> Of Lars Skjærven
> Sent: Wednesday, April 02, 2008 8:36 AM
> To: amber_at_scripps.edu
> Subject: Re: AMBER: Protein rotating out of box
>
> Hi Steve,
> Thanks for your comments. I guess the solvateoct is the way to go
> here. However, after I solvate using the command:
> 'solvateoct mol TIP3PBOX 15.0 0.78' parts of the solute is outside the
> box. Hmm.. Does not the buffer argument, in this case 15.0Å, specify
> the distance between the wall of the box and the closest atom? Or is
> it the distance to the walls in the cubic box, and then it shawes off
> the corners no matter if the solute is in the way? I have to make the

You are correct with the second point. solvateoct first creates a cubic box
using your "buffer" value and then cuts off the corners, even if that brings
the solute close to the box edge.

Hopefully others can chime in here--I don't have enough experience with
Amber to know how to solve this problem. Sorry. Maybe you can investigate
using Gromacs (or possibly VMD?) to create the solvated box for your system,
the have it write out a PDB file to input to leap. I know that Gromacs does
create a box with a "buffer" around the protein as you would expect.

> buffer as high as 33Å to get the solute entirely inside the box.
> Have a nice day.
> LarsS
>
> On Tue, Apr 1, 2008 at 4:55 PM, Steve Spronk <spronk_at_umich.edu> wrote:
> > Hi Lars,
> >
> > I know that using an octahedral box would waste waters, but this is
> > definitely the preferable method. It's better to have slower
> simulations
> > with an octahedral box than fast simulations that give you junk, like a
> > protein interacting with its own image.
> >
> > The only other method that I can think of is using a rectangular box as
> > before, but you could place very weak restraints on a couple of atoms
> that
> > lie along the long axis (preferably on opposite sides of the protein).
> That
> > way the protein couldn't rotate except around the long axis. However,
> this
> > is nonphysical and may introduce other problems--you'll have to
> investigate.
> > I've never heard of this before, but perhaps you could get away with
> it.
> >
> > People do simulations with long straight structures like DNA or fibrous
> > proteins. You could see if there are any tricks mentioned in those
> papers.
> >
> >
> > Steve
> >
> >
> >
> > -----Original Message-----
> > From: owner-amber_at_scripps.edu [mailto:owner-amber_at_scripps.edu] On
> Behalf Of
> > Lars Skjærven
> > Sent: Thursday, March 27, 2008 5:35 PM
> > To: amber_at_scripps.edu
> > Subject: AMBER: Protein rotating out of box
> >
> > Dear all,
> >
> > I am simulating a very nice little protein in a rectangular box with
> > different dimensions. Unluckily for me, the solute rotates within its
> > box during the simulation such that the longest dimension of the
> > solute no longer coincides with the longest dimension of the box.
> > Thus, it now spans the whole box, and interacting with its image. Some
> > atoms are actually out of the box on both ends. I use default values
> > of NSCM. I would avoid using a octahedral box since there will be a
> > lot of wasted waters.
> >
> > Any tips on how to tackle the rotation will be greatly appreciated.
> > Have a pleasant day :-)
> >
> > Best regards,
> > Lars Skjærven
> > Department of Biomedicine
> > Universitetet i Bergen
> > -----------------------------------------------------------------------
> > The AMBER Mail Reflector
> > To post, send mail to amber_at_scripps.edu
> > To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
> >
> >
> >
> > -----------------------------------------------------------------------
> > The AMBER Mail Reflector
> > To post, send mail to amber_at_scripps.edu
> > To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
> >
>
>
>
> --
> mvh Lars Skjærven
> Institutt for Biomedisin
> Universitetet i Bergen
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu