AMBER Archive (2009)

Subject: Re: [AMBER] Thermodynamic Integration at higher values of clambda using AMBER 9

From: steinbrt_at_rci.rutgers.edu
Date: Thu Mar 12 2009 - 19:37:08 CDT


Hi Pradeep,

it's not easy to say what is going on here. TI calculations with high
lambdas can show instabilities under certain conditions, but your input
files look fine at first glance (why do you set tautp when ntt=0? Although
thats probably not a problem). Its a bit odd that the problem occurs
>400.000 steps into the simulation.

Have you checked what your system looks like right before the crash? Does
your last printout of the energies look fine? Did you remove the methane
molecule's partial charges before disappearing it? Can you run at a
shorter timestep? Are there vlimit errors without crashes earlier in the
run? Can you run it as NVT? Can you run normal MD with both your V0 and V1
prmtop? Do the TI test cases in $amberhome/test all pass?

Kind Regards,

Thomas

> I am using Amber 9 thermodynamic integration utility to calculate
> disappearance of all atom methane from my system.
>
> I am running NPT for simulation for 2 ns for obtaining the equilibrium
> structure at every clambda followed by NVE simulations for 5 ns.
>
> The system works fine for clambda 0.00 to 0.80 (I am using a different
> technique of numerical integration instead of using the Gaussian
> quadrature
> suggested in the manual).
>
> When I run the disappearing free energy calculation at clambda 0.90 and
> higher I get the following error during NVE simulation.
>
> check COM velocity, temp: 0.000000 0.00(Removed)
> check COM velocity, temp: 0.000000 0.00(Removed)
> check COM velocity, temp: 0.000000 0.00(Removed)
> check COM velocity, temp: 0.000000 0.00(Removed)
> vlimit exceeded for step 486032; vmax = 73.1079
> vlimit exceeded for step 486033; vmax = 20.0107
> vlimit exceeded for step 486034; vmax = 21.2457
> vlimit exceeded for step 486036; vmax = 28.6820
>
> Coordinate resetting (SHAKE) cannot be accomplished,
> deviation is too large
> NITER, NIT, LL, I and J are : 0 0 2086 384 413
>
> Note: This is usually a symptom of some deeper
> problem with the energetics of the system.
>
>
> My input file for NPT simulation is as below
>
> &cntrl
> imin=0, ntx=5, irest=1, ntp=1, ntb=2,
> ntf=2, ntc=2, ntt=2, ntrx=1, ntr=0,
> temp0= 240., tempi= 240., pres0= 120.,
> taup=2.0, tautp=1.0, cut=8.0, ig=342786,
> dt=0.002, comp=44.6, icfe=1, clambda=0.90,
> nstlim=1000000, ntave=50000, ntpr=50000,
> ntwr=1000, ntwx=0, ntwe=50000, klambda=6,
> /
>
> and input file for NVE simulation is as below
>
>
> &cntrl
> imin=0, ntx=7, irest=1, ntp=0, ntb=1,
> ntf=2, ntc=2, ntt=0, ntrx=1, ntr=0,
> temp0= 240., tempi= 240., pres0= 120.,
> taup=2.0, tautp=1.0, cut=8.0, ig=342786,
> dt=0.002, comp=44.6, icfe=1, clambda=0.90,
> nstlim=2500000, ntave=250000, ntpr=250000,
> ntwr=1000, ntwx=0, ntwe=250000, klambda=6,
> /
>
> Please help me understand why this is happening. If you want to take a
> look
> at prmtop file, I would be glad to send it to you by e-mail.
>
> Thanks,
>
> Pradeep
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>

Dr. Thomas Steinbrecher
BioMaps Institute
Rutgers University
610 Taylor Rd.
Piscataway, NJ 08854

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber