I tried running this again (this time not in parallel).
Interestingly, I got significantly different results for pmemd and sander.
I run 10 steps of simulation and this is what I got (below). I checked that
..rst .in and .prmtop files are the same. The only difference was that I
ran these in two different clusers (pmemd in TSCC and sander in Harvard
Orchestra).
Step 1 using sander:
NSTEP = 1 TIME(PS) = 90.002 TEMP(K) = 299.13 PRESS =
0.0
Etot = -238642.5964 EKtot = 73013.5247 EPtot =
-311656.1211
BOND = 418.8091 ANGLE = 1103.1341 DIHED =
1624.5376
1-4 NB = 536.1276 1-4 EEL = 8158.4711 VDWAALS =
17142.4812
EELEC = -289239.4141 EHBOND = 0.0000 RESTRAINT =
0.0000
EPOLZ = -51400.2677
Dipole convergence: rms = 0.364E-04 temperature = 0.00
------------------------------------------------------------------
Step 1 using pmemd:
NSTEP = 1 TIME(PS) = 90.002 TEMP(K) = 299.04 PRESS =
0.0
Etot = -186990.7423 EKtot = 72992.1028 EPtot =
-259982.8451
BOND = 418.8091 ANGLE = 1103.1341 DIHED =
1624.5376
1-4 NB = 536.1276 1-4 EEL = 8158.4711 VDWAALS =
17142.4812
EELEC = -288966.4058 EHBOND = 0.0000 RESTRAINT =
0.0000
Ewald error estimate: 0.4460E-06
------------------------------------------------------------------
On Tue, Sep 12, 2017 at 7:25 AM, David A Case <david.case.rutgers.edu>
wrote:
> On Mon, Sep 11, 2017, Chayya Bhargac wrote:
> >
> >
> > I am trying to run a NVT simulation on AMBER.
> > When I try running the simulation with sander,
> > it doesn't run and I get an error (shown below).
> >
> > But when, I run the same using pmemd, it runs to
> > completion. What could be the reason for this?
>
> Odd. Are these parallel runs?
>
> I'll repeat the usual debugging strategy: the sander run fails on steps 4
> and
> 5. So, you need to run a short simulation (say nstlim=100), and set
> ntpr=1,
> in order to get any information on what is going on. Do the same thing for
> pmemd. See if the intial step is the same; see when the two runs diverge
> from each other.
>
> ....dac
>
>
> _______________________________________________
> 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 Sep 13 2017 - 09:30:03 PDT
