[gmx-users] How g_energy program calculates the averages ?

rashmi_chem at iitb.ac.in rashmi_chem at iitb.ac.in
Wed Mar 4 14:10:16 CET 2009


Dear gmx users,

I am running a NPT simulation of Galactose in water for 10ns using united
atom force field on gromacs 3.3.1. The mdp parameters used are:
dt                      = 0.002
nstcomm 	 	= 1
nsteps           	= 5000000 ; 10 ns (run control)
nstxout          	= 10000
nstvout 		= 10000
nstxtcout               = 250
nstenergy	        = 250
energygrps		= GLA   SOL
nstlist 		= 10
ns_type 		= grid
rlist 			= 1.1 ;(neighbor searching)
coulombtype 		= PME
;rcoulomb_switch 	= 1.0
rcoulomb 		= 1.1
vdwtype 		= shift
;rvdw_switch 		= 1.0
rvdw 			= 1.1
fourierspacing 		= 0.12 ;(nm)
ewald_rtol		= 1e-5
pme_order 		= 4  ;(electrostatics and VDW)
tcoupl         		= berendsen
tc_grps 		= GLA  SOL
tau_t 			= 0.1  0.1
ref_t 			= 300  300
pcoupl   		= berendsen
compressibility         = 4.5e-5
tau_p                   = 1.0
ref_p                   = 1
gen_vel		 	= no
constraints 		= all-bonds
constraint_algorithm 	= lincs

After 10 ns i am calculating pressure using g_energy program. Average by
g_energy comes to be arround 0.99 bars while by xmgrace it comes to be 1.4
bars.
I reran the same simulation for 10ns using nstegergy = 1, instead of 250.
this time the average pressure calculated by g_energy and xmgrace was same
i.e. 0.99 bars.
This means that g_energy calculates average using all MD steps i.e. for
pressure values written after every 0.002 ps, while the average of xmgrace
 is from pressure values mentioned in .xvg file.
I converted .edr file to a text file, it contained three columns for each
term i.e. energy, av energy and sum energy. for eg :

  time:   5.00000e-01         step:           250
               Component        Energy    Av. Energy    Sum Energy
                G96Angle   2.30488e+01   7.26315e+03   5.68798e+03
             Proper Dih.   1.11964e+01   3.20236e+03   2.63734e+03
           Improper Dih.   4.14890e+00   2.55975e+03   1.09996e+03
                   LJ-14   2.02705e+01   3.40112e+03   4.57903e+03
              Coulomb-14   4.09736e+01   1.17554e+03   1.09782e+04
                 LJ (SR)   6.84730e+03   3.30860e+06   1.69561e+06
            Coulomb (SR)  -4.08518e+04   4.90539e+06  -1.02531e+07
            Coul. recip.  -2.35144e+03   2.94449e+04  -5.88310e+05
               Potential  -3.62563e+04   2.58237e+06  -9.12079e+06
             Kinetic En.   6.61766e+03   2.14269e+06   1.66692e+06
            Total Energy  -2.96387e+04   3.00496e+05  -7.45386e+06
             Temperature   2.98376e+02   4.35590e+03   7.51579e+04
          Pressure (bar)   5.51461e+02   1.30509e+07   6.42281e+04


My question is what these numbers mean and which of these columns are used
by g_energy program to calculate the average values (say average
pressure)??

Even if nstenergy = 250 and edr file has energies written after every 0.5
ps (250 steps), still g_energy calculates average pressure based on
pressure values writen after each MD step i.e. 0.002 ps. why is this so??

Thanxs in advance.

Rashmi







More information about the gromacs.org_gmx-users mailing list