AMBER Archive (2009)

Subject: Re: [AMBER] Simulation short peptide and protein interaction

From: Carlos Simmerling (carlos.simmerling_at_gmail.com)
Date: Thu Sep 03 2009 - 08:34:27 CDT

it's very hard to tell from such runs if the parameters are not good enough,
or the simulation needs to be run longer. you need to think about what you
want to learn. if you already have the crystal structure, why are you trying
to use MD to dock it? one possibility is to use the crystal structure and
the docked form, use MD on both and calculate energetic properties to see
which may be "better" from the force field standpoint. if the crystal form
seems better, then you may need longer runs to get the docked form to
change. however, this can be very slow, and it's possible that your docked
conformation has no low-energy pathway to adopt the crystal form without
actually dissociating first- so be careful when you think that the structure
getting "worse" is a problem. maybe that's reality.

On Thu, Sep 3, 2009 at 8:22 AM, Rilei Yu <yulaomao1983_at_yahoo.com.cn> wrote:

> Dear David,
>
> Thanks for your recommendation about that paper. I have read that paper. I
> ever tried using md to refine the docked complex a short peptide (12
> residues) and its receptor. After 3 ns second, I can find the orientation
> slightly move. After supperposing it with the crystal structure (complex of
> this peptide). I find the orientation become worse that the origial docked
> one. As a short peptide, I have to run 10 or 20 ns to reach the refine
> purpose?
> Or more? Do you have such experience, if it needs more than 50 ns, we may
> cannot afford to using md to refine the complex.
> BEST WISHES,
>
> Rilei Yu
>
> --- 09年9月2日，周三, case <case_at_biomaps.rutgers.edu> 写道：
>
> 发件人: case <case_at_biomaps.rutgers.edu>
> 主题: Re: [AMBER] Simulation short peptide and protein interaction
> 收件人: "AMBER Mailing List" <amber_at_ambermd.org>
> 日期: 2009年9月2日,周三,下午8:15
>
> On Tue, Sep 01, 2009, Rilei Yu wrote:
> >
> > I want to carry out md on a small peptide (ligand, with 12 residues) and
> > protein (receptor) system. I want to simulation the interaction whole
> > process, namely, from the process of the ligand enter the pocket of the
> > protein to it completely accommodated by the receptor. Do anyone have
> > such experience? Can we use md to carry out this intention? How much
> > time it needs? As far as I know, the association rate for this short
> > peptide with its receptor is very slow. Now, I have docked the peptide
> > to the door of the pocket.
>
> I think you are being unrealistically ambitious. Even for a small, rigid
> ligand, following the equilibrium process of binding is a real challenge;
> see,
> for example:
>
> %A O. Guvench
> %A D.J. Price
> %A C.L. Brooks, III
> %T Receptor rigidity and ligand mobility in trypsin-ligand complexes
> %J Proteins
> %V 58
> %P 407-417
> %D 2004
>
> To do what you describe for a 12-residue peptide ligand is way beyond what
> is
> currently feasible with all atom models.
>
> You should ask yourself what you are hoping to learn from such an
> exercise. It may be that a coarse-grained model would be a better choice;
> or it may that you could scale back your ambitions and tackle a part of
> the overall problem at a detailed level. Start with calculations that are
> similar to those you see described in the existing literature.
>
> ...good luck...dac
>
>
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
>
>
> ___________________________________________________________
> 好玩贺卡等你发，邮箱贺卡全新上线！
> http://card.mail.cn.yahoo.com/
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber