Hi Guqin,
The behavior you see is correct. There is nothing in your simulation that is dependent on the random number stream.
ntt=1 is the Berendsen barostat - this does not use random numbers, only ntt=2 and 3 use the random number stream. I'd suggest switching to ntt=3 if you don't have a specific reason for wanting to use Berendsen. That will then get you different simulations due to different random seeds.
The other reason you don't see a dependence is because tempi=0 - Ad you noted the manual said ntx .lt. 3 AND TEMPI .ne. 0.0. You have ntx=1 but have tempi == 0.0. If you truly want different initial velocities then set tempi to a number greater than zero and you will get a boltzman distribution of velocities (based on the random number stream) to give you the initial temperature you set with tempi.
Hope that helps.
All the best
Ross
> On Oct 16, 2015, at 9:17 AM, Guqin Shi <shi.293.osu.edu> wrote:
>
> (I forgot to add [AMBER] in subject so I resend this email. Hope I didn't
> cause any confusion.. Thanks!)
>
> Hi all!
>
> recently I want to do multiple independent runs by assigning different
> velocities to each run.
>
> So I have three stages for each run, eq1 for heating up, eq2 for pressure
> relaxation, and then run at NPT conditions. The assignment of velocites is
> at eq1 stage and below is the input:
>
> *Heating up the system equilibration stage -- only water is mobile*
> * &cntrl*
> * nstlim=100000, dt=0.002, ntx=1, irest=0, ntpr=500, ntwr=5000, ntwx=500,*
> * tempi =0.0, temp0=300.0, ntt=1, tautp=2.0, ig=-1,*
> * ntb=1, ntp=0,*
> * ntc=2, ntf=2,*
> * nrespa=1,*
> * ntr=1,*
> * iwrap=1,*
> */*
> *Keep protein and ligand fixed*
> *50.0*
> *RES 1 100*
> *END*
> *END*
>
>
> However, I noticed that each run are IDENTICAL in all .out, .crd, and .rst
> files. So I realize the ig flag didn't work at all. Maybe I didn't fully
> understand the sentence in manual "The MD starting velocity is dependent on
> the random number generator seed if NTX .lt. 3 .and. TEMPI .ne. 0.0." Can
> anybody help illustrate how to correctly use ig flag? It seems ig=-1
> doesn't work if I have ntt=1 and tempi=0.0...? So in my case, what's the
> best way to assign random velocities in each run? How about assigning a
> number to ig? Do I need to change it for each run? It should be a very
> simple one but I have no experience dealing with it before.. Thanks for
> your time!
>
> Best,
> Guqin
>
> --
> Guqin SHI
> The Ohio State University
> College of Pharmacy
> 500 W. 12th Ave.
> Columbus, OH, 43210
> (614)688-3531
> _______________________________________________
> 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 Fri Oct 16 2015 - 09:30:06 PDT
