AMBER Archive (2002)

Subject: Re: Running TIP5P with PBC

From: David Case (case_at_scripps.edu)
Date: Wed Jul 03 2002 - 14:49:28 CDT

On Wed, Jul 03, 2002, Sarah Tschampel wrote:
>
> I am trying to use amber7 to run a constant pressure simulation with TIP5P
> as the solvent. I did this successfully for TIP4P solvating with the
> preequilibrated box of TIP4P. Since there is not an equivalent for TIP5P
> I am trying to obtain such a box. I have tried many things, heating over
> longer times (from 50-1000ps), trying NTT=1 and 4, as well as removing
> translational motion more that every 1000 steps. Also, I have tried to
> run a constant volume simulation first, but when I try to run constant
> pressure from the output it still fails.

What happens when it fails? Symptoms?

>
> Here is an example of one input file I have tried:
>
> heating tip5p box of water
> &cntrl
> IREST=0, IMIN=0, NMROPT=1, NTPR=5000,
> NTX=1, IG=71277, TEMPI=5.0, HEAT=5.0,
> NTB=2, NTC=2, TOL=0.0005, NTF=2, NSNB=25,
> NTT=4, TEMP0=300.0, DTEMP=5.0, TAUTP=0.25,
> NTP=2, PRES0=1.0, COMP=44.6,
> NSTLIM=300000, T=0.0, DT=0.002,
> CUT=8.0, SCNB=2.0, SCEE=1.2, DIELC=1.0,
> DX0=0.01, DXM=0.5,
> IBELLY=0, NTWX=500, NSCM = 100,
> &end
> &wt
> TYPE='TEMP0',ISTEP1=1,ISTEP2=250000,VALUE1=5.0,VALUE2=300.0,
> &end
> &wt
> TYPE='TEMP0',ISTEP1=250001,ISTEP2=300000,VALUE1=300.0,VALUE2=300.0,
> &end
> &wt TYPE='END'
> &end END
>

First, use ntp=1. Second, you have to heat the system to 300K (or
thereabouts) with constant volume, *then* turn on constant pressure.
You first weight namelist sets the target temperature to 5K, and heats
it up after that. That will almost certainly cause problems: just
run at 300K, at ntp=1.

...good luck...dac