On Wed, 9 Mar 2016 08:53:56 -0500
Arati Paudyal <apsilwal123.gmail.com> wrote:
> Sorry for all these questions again. As a beginner, I am just trying
> to connect a lot of things here and I apologize in advance if some of
> these questions sound stupid.
Your questions are not stupid but I think you should reread what Jason
has said. What you call "equilibration" is a rather arbitrary label for
the initial part of a simulation (not just heating or pressurising or
whatever your simulation setup protocol does) you would not use for
calculation of averages. How long that takes is really a system
dependent matter and not a decision you can make in advance. Likewise
for how many frames you should use. What you should really be doing is
estimate the correlation time for the property you are interested in.
Dan has suggested in another thread today that you should also look out
for convergence of the property you are interested in e.g. by comparing
the average of a later simulation period to an earlier one and repeat
that for various "blocks" of your trajectory, simple but effective. But
also plot the time series of your property and examine visually how it
is behaving.
HTH,
Hannes.
> On Tue, Mar 8, 2016 at 9:39 PM, Jason Swails <jason.swails.gmail.com>
> wrote:
>
> > On Tue, Mar 8, 2016 at 2:06 PM, Arati Paudyal
> > <apsilwal123.gmail.com> wrote:
> >
> > > Dear all,
> > >
> > > I just finished the tutorial for MMPB/GB SA with ras-raf protein
> > > and I followed the tutorial. But this time I ran the the
> > > equilibrium step for 30,000,000 steps for a total simulation time
> > > of 60 ns instead of 500 ps
> > as
> > > in tutorial. In the tutorial, there are separate production steps
> > > run
> > for a
> > > total of 2 ns. I guess, my questions is, if you run your
> > > equilibiation phase for 60 ps already, do I need to run the
> > > production step again? I checked everything including rmsd,
> > > energy and density and my system is fully equilibriated.
> > >
> >
> > There is really no difference between "equilibration" and
> > "production" in terms of simulation protocol. "Equilibration" is
> > simply what we call the portion of the production simulation in the
> > beginning that we discard because there is a high(er) likelihood
> > that it is not sampling from an equilibrium distribution.
> >
> > So if you ran 60 ns of MD, and you think that (for whatever
> > justification), the last 58 ns sample from the free energy surface
> > (i.e., the equilibrium distribution), then you should start your
> > MMPBSA.py analyses from the start of the 3rd nanosecond.
> >
> > If NOT, then when we run the mmpbsa.py script, how do we start from
> > the
> > > point where the system is already equilibriated? Is it something
> > > do do
> > with
> > > the start and end frames in the script? For eg. following is the
> > > rmsd
> > plot
> > > of my system afte 60ns of equilibriation. I can see the system is
> > > equilibriated after about 200 frames. So can I assign my start
> > > frame as
> > 200
> > > and end frame as 700?
> > >
> >
> > I would suggest setting your start frame to 200 or so, as you
> > said, and then putting your endframe to the very end of the
> > simulation. You can pick a long interval to reduce the number of
> > structures you feed to MMPBSA.py. The more time between adjacent
> > snapshots, the less correlated the results will be (and therefore,
> > the more statistically significant they will be).
> >
> > HTH,
> > Jason
> >
> > --
> > Jason M. Swails
> > BioMaPS,
> > Rutgers University
> > Postdoctoral Researcher
> > _______________________________________________
> > 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
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Mar 09 2016 - 09:00:03 PST
