AMBER Archive (2002)

Subject: ramdomized start velocities

From: Arvid Soederhaell (arvid_at_fmp-berlin.de)
Date: Fri Nov 08 2002 - 06:35:23 CST

Dear ambers

This is probably something that I should know...

I have a 6 aminoacids long peptide which i want to run a simulated
annealing on, that is simply generate lots of random conformations. I have
a script that i use to generate a series of inputfiles and then I run
these identical inputfiles. What then happens is that all structures that
are generated are identical! I thought that the start velocities in an MD
run are supposed to be random, and hence follow different trajectories...
What have i done wrong?

Here is the input:

 &cntrl
   imin=0, nmropt=1,
   ibelly=0, ntr=0,
   ntx=1, ntpr=1000, ntwx=0, ntwv=0, ntwe=0,

   ntf=2, ntc=2,
   ntb=1, ntp=0,
   dielc=1, cut=8.0, nsnb=10.0,
   scnb=2.0, scee=1.2,
   
   nstlim=25000, dt=0.001,
   temp0=300, tempi=300, ntt=4, tautp=0.2,
 &end
 &wt type='TEMP0',
    istep1=1, istep2=1000,
    value1=300, value2=300,
 &end
 &wt type='TEMP0',
    istep1=1001, istep2=5000,
    value1=300, value2=1500,
 &end
 &wt type='TEMP0',
    istep1=5001, istep2=20000,
    value1=1500, value2=1500,
 &end
 &wt type='TEMP0',
    istep1=20001, istep2=23000,
    value1=1500, value2=0,
 &end
 &wt type='END',
 &end

Arvid Soderhall