Dear Dr. Steinbrecher,
Thank you very much for your help.
1. In VMD, I loaded the prmtop file then followed by the AMBER restart (for
*rst)* or AMBER coordinate (for *crd)* accordingly. For this point, I am
checking again the options to open in VMD, something related to implicit
solvent and periodic simulation.
2. I wanted to unfold my protein using heating (from 0K to 400K) gradually
in 50ps, then MD has been running for re-folding it (temp from 400K down to
300K and keep 300K for equilibration) in ~ 60ns (it is still running now).
I want to detect the folding simulation, but because my protein is quite big
(158 aa), and I think it is infeasible to fold it from scratch (from its
sequence), so I started with a process of unfolding a homology model, and
make it re-fold again.
3. I want to visualize it because I want to make sure it is unfolded before
running large amount of time to refold it, then it might trap in some local
minima (It did happen actually, the RMS to the average structure was stable
in some short MD run ~ 6ns, and its deviation to the crystal structure is
1.9 angstrom. However, I was told that this might be trapped in some local
minima.)
I hope my reasoning is not so wrong so far. As you said, it was NOT
unfolded, so will you suggest me to increase the temperature to heat it up
OR could you give me some suggestion to unfold it please?
Thank you very much for your help. It really helps.
Regards,
Chinsu
On Mon, Sep 19, 2011 at 3:38 PM, <steinbrt.rci.rutgers.edu> wrote:
> Hi,
>
> > I want to unfold my protein by heating it up to 400K, then following by
> MD
> > run to expect it re-fold again. I got some results (png attached), but Im
> > not sure that it was unfolded with that temperature.
> > I viewed the results using VMD, but I got confused with the differences
> of
> > using *rst* and *crd* files.
> > The *rst* showed me the "full" structure while the *crd *showed me the
> > "scattered" structure, especially in the MD frames.
>
> The pictures from the crd files look more like representation errors, you
> do not say how you loaded the structures in vmd, but if the final crd
> snapshot and the final rst look considerably different, something is wrong
> (and it might not be Amber). VMD has several options about input formats
> and if you specify the wrong one, it still tries very hard to load the
> coordinates anyway, often leading to strange looks.
>
> You dont specify how long you simulated, but from the rst files, I would
> say your protein is definitely not unfolded. My guess would be that even
> if you manage to completely unfold your protein, there is no good reason
> to expect it to quickly refold during another simulation, but your
> literature sources may tell you otherwise for this special case...
>
> I wouldnt define folded and unfolded by just looking at the structure,
> instead rmsd-values, number of Hbonds, trajectory clustering or radius of
> gyration measurements over the trajectory should give you a lot more solid
> information...
>
> > Btw, I found a related issue in AMBER forum, and Jason said that "*The
> > restart is actually a full time step ahead with the coordinates (and a
> > **half
> > time step ahead for velocities)", may I have it more clearer?
>
> Check e.g. Allen, Tildesley, "Computer simulations of liquids", 1987
> Clarendon Press, or any other textbook that explains the different types
> of verlet propagators. This is not related to your unfolding question,
> though.
>
> Kind Regards,
>
> Thomas
>
> Dr. Thomas Steinbrecher
> formerly at the
> BioMaps Institute
> Rutgers University
> 610 Taylor Rd.
> Piscataway, NJ 08854
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Sep 19 2011 - 03:30:02 PDT
