From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Thu May 25 2006 - 09:56:47 CDT
you are correct that there is an energy loss. This is unlikely to "mess
up" the simulation, but it does lead to a temporarily unphysical state
(although, having the new terminii in place plus a bond/constraint there
is also an unphysical state ;-). If you're running with a thermostat
this is unlikely to cause big problems. If you want to carry out the
transition more smoothly, you might want to have a look at namd's FEP
which would, for example, allow you to transition between the fully
bonded state and the unbonded state with appropriate terminal patches.
patrick wintrode wrote:
> Thanks for the reply.
> A question:
> There's a certain amount of energy associated with a covalent bond. If
> the bond suddenly disappears, won't the instantaneous jump in energy
> mess up the simulation? (This was my rational for using constraints to
> mimic the presence of a bond instead).
> Patrick L. Wintrode
> Assistant Professor
> Department of Physiology & Biophysics
> Case Western Reserve University
> Cleveland, OH 44106
> */Peter Freddolino <petefred_at_ks.uiuc.edu>/* wrote:
> Hi Patrick,
> for 1), we probably need a more detailed description of the error you
> get. Normally VMD plays nicely with preparing psfgen/namd input, but
> there's no reason I can think of that SPDBV wouldn't work.
> For 2), what you might just want to do is prepare 2 psf files, one of
> which actually does have a bond there. There's no technical
> problem in
> switching psf files between simulation sas long as they have they
> number of atoms.
> patrick wintrode wrote:
> > Hi all.
> > I'm trying to set up a complicated simulation. Essentially, I
> want to
> > use NAMD to look at how a protein responds when one of its loops is
> > cleaved by a protease. My plan is to remove the peptide bond in the
> > original pdb file (adding appropriate patches to the new N and C
> > termini), then use restraints in order to mimic the presence of the
> > bond. After the system has equilibrated, I'll turn off the
> > and watch how the system evolves in response.
> > My questions:
> > 1) what's the best way to prepare the modified pdb file? I tried
> > it with SwissPDB, but when I used the resulting modified PDB
> file with
> > PSFGEN, the new cleaved ends caused a fatal error.
> > 2) does anyone know of the best place to look for how to set the
> > restraints to mimic the presence of a peptide bond as accurately as
> > possible?
> > Thanks.
> > Patrick L. Wintrode
> > Assistant Professor
> > Department of Physiology & Biophysics
> > Case Western Reserve University
> > Cleveland, OH 44106
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:43:38 CST