Steered Molecular Dynamics to compute protein-protein interaction energies

[joseph]

Dear PLUMED Users,

I'm using gromacs 4.5.1 with amber force-field to study a protein-
protein complex (roughly 600 and 20 amino acids each).
I would like to compute de binding free energies using SMD or umbrella
sampling by pulling the peptide away from the protein.

I've read Lemkul and Bevan article ( http://pubs.acs.org/doi/abs/10.1021/jp9110794
). They conducted umbrella sampling on 31 starting points during 10ns.
I would like to follow their protocol however, it seems unfeasible
with the computationnal power that is accessible to me.

Do you think that I can obtain comparable results with STEER using
fewer simulation runs (at different KAPPA values)?

Thank you very much

Joseph

[Giovanni Bussi]

Dear Joseph,

for such a complex system I think that a total of 300 ns is definitely
a lower bound for the required computational resources. Unless the
binding is trivial (which I think it will not in your case), or you
find a manner to describe the separation using "perfect" collective
variables (which could be quite difficult), there will be no free
lunch for you...

Anyway, for a *very qualitative* insight, you can try to do a single,
as slow as you can afford, STEERed MD detaching the two proteins. But
forget about obtaining a converged free energy landscape.

Best,

Giovanni

> --
> You received this message because you are subscribed to the Google Groups "PLUMED users" group.
> To post to this group, send email to plumed...@googlegroups.com.
> To unsubscribe from this group, send email to plumed-users...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/plumed-users?hl=en.
>
>

[joseph]

Dear Giovanni,

thank you for your insight. I'll think about it.
In the mean time I'll also consider MM-PBSA as I supposed it would be
a reasonable approach for ma problem.

Best,

Joseph

On 14 avr, 09:59, Giovanni Bussi <giovanni.bu...@gmail.com> wrote:
> Dear Joseph,
>
> for such a complex system I think that a total of 300 ns is definitely
> a lower bound for the required computational resources. Unless the
> binding is trivial (which I think it will not in your case), or you
> find a manner to describe the separation using "perfect" collective
> variables (which could be quite difficult), there will be no free
> lunch for you...
>
> Anyway, for a *very qualitative* insight, you can try to do a single,
> as slow as you can afford, STEERed MD detaching the two proteins. But
> forget about obtaining a converged free energy landscape.
>
> Best,
>
> Giovanni
>
>
>
>
>
>
>
> On Wed, Apr 13, 2011 at 6:26 PM, joseph <andrejos...@gmail.com> wrote:
> > Dear PLUMED Users,
>
> > I'm using gromacs 4.5.1 with amber force-field to study a protein-
> > protein complex (roughly 600 and 20 amino acids each).
> > I would like to compute de binding free energies using SMD or umbrella
> > sampling by pulling the peptide away from the protein.
>

> > I've read Lemkul and Bevan article (http://pubs.acs.org/doi/abs/10.1021/jp9110794