[gmx-users] Problem with Simulated Annealing
Gaurav Porwal
gauravp at che.iitb.ac.in
Sat Feb 28 12:29:01 CET 2004
Dear Gromacs users,
I'm facing a very strange kind of problem.
Currently, I'm working on minimising the energy of the structures using
simulated annealing to get globally minimum energy structure and so using
GROMACS for the purpose.
I've generated about 1000 random protein structures for a 20 residue
protein using Ribosome. Using a shell script I input each structure and
minimise the energy and then store the minimised structure in a separate
directory.
In the script, first I'm minimising energy using 'md' integrator without
simulated annealing at 300 K, using the file vac.mdp given below. After
this I again run mdrun, now with Simulated Annealing activated from 800K
to 300K, using the file cool.mdp also given below. The shell script is
also given below.
The problem I'm facing after this minimisation is that in the random
structure there are some secondary structures eg.helix, but in the
minimised structure, all the secondary structures vanish which is a kind
of wierd as occurence of secondary structure generally signifies lower
energy. Also not much bending of the back bone occurs, as in, no closed
structure is formed.
Can somebody plz. enlighten why this is happening and how it can be
improved.
The relevant files are as follows:
Shell Script
------------
cp ../pdb3/$1.pdb .
echo 3|pdb2gmx -f $1.pdb -o $1.gro -p $1.top
editconf -f $1.gro -o b4vac.gro -box 10 10 10
grompp -v -f vac.mdp -c b4vac.gro -o vac.tpr -p $1.top
mdrun -v -s vac.tpr -o vac.trr -c after_vac.gro -e vac.edr -g vaclog
grompp -v -f cool.mdp -c after_vac.gro -o vac.tpr -p $1.top
mdrun -v -s vac.tpr -o vac.trr -c after_vac1.gro -e vac.edr -g vaclog
editconf -f after_vac1.gro -o $1m.pdb
g_rms -s vac.tpr -f vac.trr -o $1rmsd.xvg -m $1rmsd.xpm -bin $1rmsd.dat
cp $1m.pdb ../pdb3/.
cp $1rmsd.xvg ../pdb3/.
rm -f \#*
rm -f *.gro *.top *.itp *.pdb *.tpr *.trr *.log mdout.mdp *.edr *.dat
*.xvg *.xpm *.eps
where $1 is for the imcoming random file kept in directory pdb3. We are
using the option 3 in the forcefield selection
vac.mdp
-------
;
; User spoel (236)
; Wed Nov 3 17:12:44 1993
; Input file
;
cpp = /lib/cpp
define = -DFLEX_SPC
constraints = none
integrator = steep
nsteps = 1000
;
; Energy minimizing stuff
;
emtol = 2000
emstep = 0.01
nstcomm = 1
ns_type = grid
rlist = 1
rcoulomb = 1.0
rvdw = 1.0
Tcoupl = yes
Pcoupl = no
gen_vel = no
annealing = no
zero_temp_time = 0
tc-grps = Protein
tau_t = 0.1
ref_t = 300
cool.mdp
--------
title = 30mer
cpp = /lib/cpp
constraints = all-bonds
constraint_algorithm = lincs
integrator = md
tinit = 0.0
dt = 0.001 ; ps !
nsteps = 5000 ;
nstcomm = -1
nstxout = 50000 ;
nstvout = 50000 ;
nstfout = 0
nstenergy = 50000 ;
nstlog = 50000
energygrps = protein
nstlist = 10
ns_type = simple
epsilon_r = 1
coulombtype = shift
vdwtype = shift
rlist = 1.4 ; cut-off for ns
rvdw = 1.4
rvdw_switch = 0.8
rcoulomb = 1.4
rcoulomb_switch = 0.8
epsilon-r = 1.0
; Temperature coupling is on in two groups
Tcoupl = yes
tau_t = 0.1
tc-grps = protein
ref_t = 800
; Pressure coupling is not on
Pcoupl = no
tau_p = 0.5
compressibility = 4.5e-5
ref_p = 1.0
; Generate velocites is on at 800 K.
gen_vel = yes
gen_temp = 800.0
gen_seed = 173529
annealing = yes
zero_temp_time = 800000
disre = simple
disre_weighting = equal
----------------------------------------------------
Thanks in advance.
Regards,
Gaurav
*********************************************************************
I can't believe that God put us on this earth to be ordinary.
---- Lou Holtz
*********************************************************************
Gaurav Porwal
Senior Undergraduate Student
Department of Chemical Engineering
IIT Bombay, Powai
Mumbai - 400076
INDIA
Ph. - 9820365472
Alternative Email Id's :-
porwal at iitb.ac.in
porwal82 at yahoo.com
**********************************************************************
More information about the gromacs.org_gmx-users
mailing list