On Thu, Aug 13, 2015 at 8:36 PM, Amber mail <amber.auc14.gmail.com> wrote:
> Dear AMBER users,
>
> I am performing a MD simulation using AMBER12 under the ff99SB force filed.
> Initially, the structure was neutralized and then solvated using TIP3P
> water model.
>
> After the minimization stage, a 50ps of MD simulation was performed from 0K
> to 100K. The system is then heated up in increments of 25K with 50ps of MD
> simulation at each temperature increment until the desired temperature of
> 310K was established.
>
> This the control file for heating (0-100)K
>
> Heat
> > &cntrl
> > imin=0,
> > ntx=1,
> > irest=0,
> > nstlim=25000,
> > dt=0.002,
> > ntf=2,
> > ntc=2,
> > tempi=0.0,
> > temp0=100.0,
> > ntpr=100,
> > ntwx=100,
> > cut=8.0,
> > ntb=2,
> > ntp=1,
> > ntt=3,
> > gamma_ln=1.0,
> > tautp=1.0,
> > taup=1.0,
> > pres0=1.01325,
> > nmropt=0,
> > ig=-1,
> > iwrap=1,
> > /
> >
>
> A 50 ns of MD simulation is going to be performed at 310K. Till now, I have
> got 20 ns out of 50 ns. This is the control file for the production step
>
> Production
> > &cntrl
> > imin=0,
> > ntx=5,
> > irest=1,
> > nstlim=5000000,
> > dt=0.002,
> > ntf=2,
> > ntc=2,
> > temp0=310.0,
> > tempi=310.0,
> > ntpr=5000,
> > ntwx=5000,
> > cut=8.0,
> > ntb=2,
> > ntp=1,
> > ntt=3,
> > gamma_ln=1.0,
> > ig=-1,
>
> nmropt=0,
> > iwrap=1,
> > /
> >
>
> The problem is that a part of the protein is leaving the water box after 20
> ns of the MD production
>
>
>
> Please correct me If I am wrong, since I used the wrapping option
> (iwrap=1), the water molecules should wrap (surround) any residue in case
> if it is trying to leave the water box. does this mean that these outer
> residues are being simulated now in vacuum?!
No, this is not what it means. Your simulation is still being simulated
under periodic boundary conditions, which means that *all of space* is
filled with copies of particles translated by whole periodic box vectors.
Consider a basic cartoon of periodic boundary conditions (e.g.,
http://dynamomd.org/images/PBC.png). Note that in that image, there are an
*infinite* number of choices you can make about how to define the "primary
box". You can translate it arbitrarily in both dimensions (all three
dimensions if you have a 3-D periodic system, like in MD simulations) so
that it cuts through the middle of any of the circles.
One way to define the periodic cell is to put the protein in the center and
all of the water around it. That is probably what you would most like to
see, but that's no more "correct" from a simulation perspective than
translating the box so that the center of mass of the protein is next to
one of the edges (which means that part of the protein appears to "stick
out" of the primary unit cell). If you want a different view, you need to
image your trajectory to achieve it. For 99% of applications, the
"autoimage" command in cpptraj will give you the "prettiest" unit cell
representation.
> If I am right, is setting
> ntb=2 (boundary conditions) works for this behavior, and there is no
> problem ?!
>
No. ntb=2 simply means "use periodic boundary conditions, but let the
unit cell change size and maybe shape". This happens when you run constant
pressure simulations. Periodic boundary conditions are periodic boundary
conditions, whether you set ntb=1 or ntb=2, there's no difference in this
regard.
> Another question regarding the heating stage, below is the plot of
> Temperature vs. Time before production, I am wondering why the increase in
> the Temperature was not going smoothly and the equilibration did not reach
> at the end of the heating stage (like the results of the tutorials, which I
> performed before)
>
>
> A same pattern (sudden change) was also obtained for the Potential energy,
> Kinetic Energy and the Total Energy (in the heating stage)
>
That is because the heating occurs rapidly at the start of each stage.
Generally this isn't a huge problem, but you can use nmropt=1 with
temperature control to slowly vary the target temperature in order to heat
the system steadily from low to high temperature in a single simulation.
There are examples (called "slow heat") in my input file repository at
https://github.com/swails/Mdins
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Aug 14 2015 - 05:00:03 PDT
