On Wed, Apr 29, 2015 at 3:46 AM, Wang Moye <wangmoye13.mails.ucas.ac.cn>
wrote:
>
>
>
>
> Hello everyone,
>
> I want to consult you some questions. I have a REMD process using AMBER12.
> In my input files ( one of the input file is shown blow ), I use Langevin
> dynamics by setting ntt=3 , and use implicit water models by setting igb =2.
>
> The input file for one replica in REMD:
>
> &cntrl
>
> irest=0, ntx=1,
>
> nstlim=2000, dt=0.002,
>
> irest=0, ntt=3, gamma_ln=1.0,
>
> temp0=260, ig=-7039,
>
> ntc=2, ntf=2, nscm=1000,
>
> ntb=0, igb=2,
>
> cut=999.0, rgbmax=999.0,
>
> ntpr=1000, ntwx=1000, ntwr=1000,
>
> numexchg=1000,
>
> /
>
> &wt TYPE='END'
>
> /
>
> (1) Can the Langevin dynamics and implicit water models be used in the
> same REMD process?
>
Yes (and is in fact recommended).
>
> (2) When I am using Langevin dynamics, how to choose the ig value?
>
My general rule -- always set it to -1. This will make the random number
generator seed from the wall clock (so it will be different for every
replica every time). Otherwise, just make sure it is different for each
mdin file each time you restart the simulation.
>
> (3) Is it necessary to change the ig value when the simulations be
> restarted?
>
Only if you choose a value besides -1.
>
> (4) Can the Langevin dynamics be repeated totally? Or can’t?
>
I don't know what this means?
> As for my progress, the jobs must be frequently restarted ( one of the
> restart input files is shown below).
>
> The restart input file for one replica in REMD:
>
> &cntrl
>
> irest=1, ntx=5,
>
> nstlim=2000, dt=0.002,
>
> irest=1, ntt=3, gamma_ln=1.0,
>
> temp0=260, ig=4000,
>
As suggested above, use ig=-1
>
> ntc=2, ntf=2, nscm=1000,
>
> ntb=0, igb=2,
>
> cut=999.0, rgbmax=999.0,
>
> ntpr=1000, ntwx=1000, ntwr=500000,
>
> numexchg=2000,
>
> /
>
> &wt TYPE='END'
>
> /
>
> (5) Which parameter should be modified to make sure restart normaly?
>
> I modify the irest =1, ntx=5, but the restart trajectories are wrote in
> the wrong way.
>
What does this mean, "the restart trajectories are wrote in the wrong
way"? How was it "broken"? What did you expect it to have?
>
> (6)Should I modify the temp0 and ig at every restart time? And how?
>
Again, set ig=-1 and forget about it. temp0 never has to be changed. The
temperature for each replica will be set by the restart file if the
temperature is stored there (it will override the temp0 from the mdin
file), which is the the correct way to restart a REMD simulation. So in
general, you never have to adjust mdin files when restarting any kind of
REMD simulation in Amber.
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 Wed Apr 29 2015 - 05:30:02 PDT
