AMBER Archive (2009)
Subject: Re: [AMBER] (no subject)
From: Carlos Simmerling (carlos.simmerling_at_gmail.com)
Date: Fri Jan 16 2009 - 12:46:11 CST
only the bonds that you added in leap should be long. you don't say
how you built your initial coordinates so it is hard to help.
On Fri, Jan 16, 2009 at 1:42 PM, Jenny Iskrenova <jnova2001_at_gmail.com> wrote:
> I have the same problem. Although I start with a good structure. The
> structure was created and the geometry was optimized in Materials
> Studio. I load the good initial pdb file in LeAP and add ions. Then I
> create the inpcrd and prmtop files. When viewed with VMD, the
> structure in the inpcrd file is distorted -- some of the bonds are
> really stretched.
> After solvating, I perform equilibration runs. Then, I proceed with
> minimization runs. At the end of all this initial equilibration and
> minimization, the structure looks reasonable. I haven't done the
> "bring the atoms slowly together" trick with nmropt=1. There are not
> just two atoms that are too far apart. My molecule is an acetate ion.
> I suppose, for such a small molecule, one can afford longer
> equilibration/minimization runs. But is this enough?
> On Fri, Jan 16, 2009 at 12:43 PM, Carlos Simmerling
> <carlos.simmerling_at_gmail.com> wrote:
>> leap can add the bond, but that doesn't automatically change the
>> peptide conformation to make the bond short. all leap will do is take
>> the conformation that you loaded or built, and create a bond. you will
>> need to minimize this or do some work in building your initial
>> conformation to get it right.
>> note that minimizing a long bond is VERY likely to cause other
>> problems, since the energy will eb high enough to invert chiralities,
>> flip amides and so on.
>> I suggest that if you don't have a good conformation to start with,
>> that you build it without the cyclic bond, and then create a distance
>> restraint in sander for these 2 atoms. then, use nmropt=1 to slowly
>> draw them together during MD, with a final distance the same or close
>> to the bond length. take the final structure, convert to pdb, and use
>> that as the initial structure for the prmtop that has the bond
>> On Fri, Jan 16, 2009 at 12:38 PM, Lake, Thomas
>> <thomas.lake08_at_imperial.ac.uk> wrote:
>>> I am trying to model a cyclic peptide, so I loaded the pdb into LEAP. I
>>> then created a bond between the two terminal residues, and then altered
>>> the partial charges and the atom types of these residues. In LEAP this
>>> seemed to work fine; I then created topology and coordinate files using
>>> saveamberparm. However when I loaded these into VMD, the bonds where
>>> the termini were linked are now massively distorted. In the LEAP editor
>>> there doesn't seem to be this problem. Can anyone help me.
>>> AMBER mailing list
>> AMBER mailing list
> AMBER mailing list
AMBER mailing list