AMBER Archive (2006)

Subject: Re: AMBER: LES copies

From: Carlos Simmerling (carlos_at_csb.sunysb.edu)
Date: Mon Apr 24 2006 - 06:14:20 CDT


since the copies have the same initial coordinates and the same
forces, and you assigned them identical initial velocities (from tempi=0)
they will not move apart. you need tempi>0.
This is discussed in the Amber manual- for Amber8 it is on page
186 in the paragraph starting "IMPORTANT"

also my personal feeling is that heating from temperature 0 to 300
can be very dangerous- nearly infinite scaling of the velocities so if there
is any problem in the initial structure it can be magnified too much.
I prefer to assign tempi=100 or something similar, then the scaling is
only 3x.
carlos

Claudia Steinert wrote:

>Hello Amber Users,
>i tried to copy a region in my protein with amber LES.
>My protein consists of 642 atoms (46 aa) and i wanted to copy atom no 230
>to 346 five times.
>I used the following input for addles:
>
>file rprm name=(crm_wat.top) read
>file rcrd name=(crm_min_wat.crd) read
>file wprm name=(crm_addles_wat.top) wovr
>file wcrd name=(crm_addles_wat.crd) wovr
>action
>spac numc=5 pick #prt 230 346 done
>*EOD
>
>and this one for sander_les:
>
># md using sander heatup tempi=0 isotrop press scaling
> &cntrl
> irest=0, ntx=1, imin=0, nstlim=150000,
> nsnb=25, tempi=0.0, temp0=300.0, scee=1.2,
> ntt=1, tautp=1.0, dt=0.001,
> ntc=2, ntf=2, cut=10.0,
> ntb=2, ntp=1, nmropt=0,
> ntpr=100, ntwe=100, ntwx=100,
> ntr=0,
> /
>
>after using addles i created a pdb file and checked, that the copies were
>actually made.
>
>as i tried to extract the different copies after sander_les in moilview,
>there was no difference between the resulting trajectories and
>structures.
>
>even as i increased the temperature in sander_les the trajectories
>remained identical.
>
>does anybody know, where the problem is?
>
>thank you in advance,
>claudia
>
>
>
>
>
>
>

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