AMBER Archive (2004)

Subject: Re: AMBER: how to correct the charge of oligosaccharides

From: Karl N. Kirschner (crux5_11_at_yahoo.com)
Date: Tue Aug 10 2004 - 09:53:29 CDT

Hi Wen-Chi,
  Below are some comments that I hope help you in your quest to build
oligosaccharides.

> > x = sequence {OME 1GA 4GA}
> > y = sequence {OME 1GA 4GA}
> > z = sequence {OME 1GA 4GA 4GA}

This is the incorrect format for building a oligosaccharides. The 1GA
is a terminal sugar and the 4GA is a linking sugar (it is linked at the
C1 and the O4 position). The proper format is
x = sequence {OME 4GA 1GA}
This is outlined in the glycam tutorial distributed with amber8, as well
as in the manual. This will also correct the message about concerning
missing parameters.

> > check x
> Checking 'x'....
> Warning: Close contact of 0.755725 angstroms between .R<OME 1>.A<
> ...
> check: Warnings: 3
> Unit is OK.
> > check y
> Checking 'y'....
> Warning: Close contact of 0.755725 angstroms between .R<OME 1>.A<

These are just warnings - they should not be a problem is you do a
minimization to clean up the structure. If you are uncomfortable with
the warnings, you can impose torsion values to generate a better initial
structure during the leap build by using the impose command.

> > charge x
> Total unperturbed charge: 0.0050
> Total perturbed charge: 0.0050
> > charge y
> Total unperturbed charge: 0.0080
> Total perturbed charge: 0.0080
> > charge z
> Total unperturbed charge: 0.0110
> Total perturbed charge: 0.0110
>

These charge errors are a result of rounding.
4GA should have a zero charge, OME should have a -0.194, and 1GA should
have +0.194. To get a perfectly neutral system, alter the charge on 1GA
C1 and the 4GA C4 appropriately.

>
> I also saved these oligos in pdb file, and used vmd, swiss-pdbViewer,
> and xleap to visualize the molecules. These structures could be seen
> correctly only under xleap. Under vmd and spdbv, incorrect linkages
> were found. Does anyone know how to fix this?

Xleap connects the atoms based on the prep file, which is why you see
the "correct" structure. The other programs connect atoms with bonds
based on atom proximity, so if two atom are close in distance a bond
will be formed between them.

Hope this helps.
Karl 

-- 
---------------------------------
Karl Nicholas Kirschner, Ph.D.
Assistant Professor of Chemistry
Hamilton College, NY 13323
---------------------------------

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