AMBER Archive (2009)

Subject: RE: [AMBER] Maxwell-Boltzmann distribution and ig=-1

From: Ross Walker (
Date: Tue Mar 31 2009 - 22:52:42 CDT

> L. S. Caves, J. D. Evanseck, and M. Karplus, “Locally accessible
> conformations of proteins: Multiple molecular dynamics simulations of
> crambin”,* *Protein Science* *(1998), 7649-666.
> In the last paper (which is old) they advocate using multiple short
> simulations with new velocity assigments. I think the only way to do this
> in
> AMBER is by setting ntx=0, irest=1, and ig=-1 otherwise the same
> velocities
> will be assigned for all simulations. Some results have interesting
> features
> and I was wondering what seeds were used, and hence what initial
> velocities
> were being assigned.

IG is unfortunately only written if ntt=2 or ntt=3 even if you have ntt=0
and are assigning velocities at a finite temperature. Hence the seed that
was used is missing from your output files unfortunately and there is really
no way to recover what it was. However, for future runs you could edit
mdread.f as follows:

Line 979:
      else if( ntt == 4 ) then
         write(6,'(/a)') 'Nose-Hoover chains'
         write(6,'(5x,(a,f10.5))') 'gamma_ln=', gamma_ln
         write(6,'(5x,(a,i8))') 'number of oscillators=', nchain
      else if ( ntt == 0 .and. irest == 0 .and. temp0 > 0.0d0 ) then
         write(6,'(/a)') 'Initial Temperature Generation:'
         write(6,'(5x,4(a,i8))') 'ig =',ig
      end if

Good luck,

|\oss Walker

| Assistant Research Professor |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- |
| | PGP Key available on request |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.

AMBER mailing list