AMBER Archive (2005)

Subject: Re: AMBER: SHAKE error

From: Carlos Simmerling (carlos.simmerling_at_stonybrook.edu)
Date: Fri Jul 22 2005 - 09:56:53 CDT

I'm not sure what the problem is, but you might want to remove center
of mass motion in the system (nscm=500 or so).
You might want tighter T coupling during equilibration (tautp=0.1)
since your T is very high. Also a smaller pressure coupling constant
during that part, like 0.1 . Maybe try 1fs time step during
equilibration too.
something is going wrong that the T is so high. What is the T/energy
at the start of this equilibration (after the first run)?

I'm not sure that I would call your first run "heating" since it assigns
300K as the target at the beginning- it is very fast heating.
Explicit solvent runs often need a more complex equilibration than you
used, such as restraining the solute and running MD only on the
water, etc. You might check out the tutorials for examples if you
haven't done that kind of thing already.

Agata Jurkiewicz wrote:

>Dear Amber users,
>
>I run a MD simulation for a small molecule (dUMP) in TIP3P water.
>I've tried waterboxes of 6 and 10 A.
>I minimized the system before MD to get rid of bad contacts (there
>weren't any, anyway).
>Then I run 50 ps of MD to heat the system to 300 K. The md.in file
>goes like this:
>
>Amber 8, heating
> &cntrl
> imin=0, irest=0, ntx=1, ntp=0, nstlim= 25000,
> ntb=1, dt=0.002, ntt=1, temp0=300, tempi=0.1,
> ntpr=50, ntwr=500, scee=1.2, ntf=2, ntc=2,
> cut=12, ntwe=500, ntwx=500, ntave=5000, tautp=2.5,
> /
>
>and the simulation goes OK.
>
>Then I try to equilibrate the system at constant pressure.
>The md.in file for equlibration is:
>Amber 8, eqlbrtn
> &cntrl
> imin=0, irest=1, ntx=7, ntp=1, nstlim= 25000,
> ntb=2, dt=0.002, ntt=1, temp0=300, tempi=300,
> ntpr=50, ntwr=500, scee=1.2, ntf=2, ntc=2,
> cut=12, ntwe=500, ntwx=500, ntave=5000, tautp=2.5,
> /
>
>It goes about 12-17 ps (depending on the size of waterbox) and fails
>with a message:
>[from md.out file:]
>
>NSTEP = 6300 TIME(PS) = 62.600 TEMP(K) = 479.96 PRESS = 1029.9
> Etot = -3922.0613 EKtot = 2086.3751 EPtot = -6008.4364
> BOND = 8.2350 ANGLE = 16.6152 DIHED = 17.7050
> 1-4 NB = 7.1684 1-4 EEL = -107.8100 VDWAALS = 817.9307
> EELEC = -6768.2806 EHBOND = 0.0000 RESTRAINT = 0.0000
> EKCMT = 1007.0906 VIRIAL = 403.0253 VOLUME = 27165.7778
> Density = 0.8072
> Ewald error estimate: 0.1265E-02
>
>------------------------------------------------------------------------------
>
> vlimit exceeded for step 6301 ; vmax = 103.896044101908
>
> Coordinate resetting (SHAKE) was not accomplished
> within 3000 iterations
>
> Note: This is usually a symptom of some deeper
> problem with the energetics of the system.
>
>I viewed all the trajectories in VMD and didn't see anything unusual
>in the geometry of the system.
>I tried this calculation on a different machine, with the same result.
>I've used the same input files to run MD for a protein-ligand system,
>with the same ligand, successfully.
>I will appreciate any hints.
>
>Regards,
>Agata.
>
>
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu