Does your Arg residue have any initial momentum in the Y-dimension to start
with (you'd have to calculate that from the velocities in the restart
file). What about the whole simulation cell? I can imagine that by
changing the masses of only a portion of the system, you incidentally
imparted net momentum in the Y-dimension for the unit cell as a whole.
Perhaps you should try setting ntx=1, irest=0 so that the velocities are
resampled from a Maxwell-Boltzmann distribution (which should have no net
momentum, or at least very little of it). Does this fix your systematic
drift in the Y-dimension?
I'll also point out that thermostatted diffusion is not "real" diffusion in
that it cannot be compared to experimental measurements (you need to run
NVE with good energy conservation to achieve that).
HTH,
Jason
On Mon, May 6, 2013 at 9:25 AM, Deák Robert <kokumetto.gmail.com> wrote:
> Sorry, the input file is here.
>
>
>
> 2013/5/6 Deák Robert <kokumetto.gmail.com>
>
> > Dear Amber users,
> >
> > Recently I have done a test.
> >
> > The system is a single ARG solvated in a truncated octahedron water box,
> > with a 9A TIP3P water buffer aroud the ARG.
> > A have minimized, heated to 340 K and equlibrated the system, like in the
> > Tutotial 1, and the production MD (the input file prod.in is attached)
> > went well.
> >
> > Now comes the test.
> > I have multiplied all the masses in the system with 4 (using parmed). The
> > new dt and taup are doubled and gamma_ln is the half of the earlier one.
> > Theoretically the extracted data for the diffusion of the ARG and the
> > increased mass ARG (mass x 4) are the same.
> >
> > While for the first system everything seems OK, following the center of
> > mass of ARG (mass x 4) the y coordinate of the molecule is almost
> > continously increasing, while the x and z shows diffusion like paths.
> >
> > Can somebody explain me why is happening this, or this is a bug?
> >
> > Regards,
> > Robert
> >
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon May 06 2013 - 07:30:02 PDT
