Project

General

Profile

Bug #464

g_velacc in gromacs-4.0.7 seems to be producing some erroneous result.

Added by Vasudevan Venkateshwaran about 9 years ago. Updated about 9 years ago.

Status:
Closed
Priority:
Normal
Assignee:
Erik Lindahl
Category:
analysis tools
Target version:
Affected version - extra info:
Affected version:
Difficulty:
uncategorized
Close

Description

The velocity auto correlation function was calculated using g_velacc for a waterbox containg 213 spc/e water molecules.

the command used:

gromacs 4.0.7 -
g_velacc_v4 -f traj.trr -s topol.tpr -n index.ndx -o vac-gmx4.xvg

gromacs 3.3.3 -
g_velacc_v3 -f traj.trr -s topol.tpr -n index.ndx -o vac-gmx3.xvg

In both cases SOL was the group selected to do the calculation. The result on gromacs 4 was different from the one in gromacs 3. the following mdp options were used for the simulation

run.mdp
---------
dt = 0.001
nsteps = 50000
comm-mode = Linear
nstcomm = 1
comm-grps = system
nstxout = 1
nstvout = 1
nstfout = 0
nstcheckpoint = 1000
nstlog = 1000
nstenergy = 1
nstxtcout = 1000
xtc-precision = 1000
xtc-grps = system
energygrps = system
nstlist = 5
ns-type = grid
pbc = xyz
rlist = 0.9
domain-decomposition = no
coulombtype = PME
rcoulomb-switch = 0
rcoulomb = 0.9
epsilon-r = 1
vdw-type = Cut-off
rvdw-switch = 0
rvdw = 0.9
table-extension = 1
tcoupl = Berendsen (V-rescale was used for the gmx4 version)
tc-grps = system
tau-t = 0.1
ref-t = 300
Pcoupl = No
Pcoupltype = Isotropic
tau-p = 1
compressibility = 4.5e-5
ref-p = 1
gen-vel = yes
gen-temp = 300
gen-seed = 173529
constraint-algorithm = Lincs

Attached is the graph comparing the output of gromacs 3.3.3 and gromacs 4.0.7

figs.png (12.7 KB) figs.png plot comparing result from gromacs 3 and gromacs 4 Vasudevan Venkateshwaran, 07/10/2010 06:04 PM
vac3.3.3.xvg (562 KB) vac3.3.3.xvg VACF for SPC/E with gromacs 3.3.3 David van der Spoel, 07/13/2010 01:28 PM

History

#1 Updated by Vasudevan Venkateshwaran about 9 years ago

Created an attachment (id=488)
plot comparing result from gromacs 3 and gromacs 4

#2 Updated by David van der Spoel about 9 years ago

Created an attachment (id=490)
VACF for SPC/E with gromacs 3.3.3

I get the same VACF with gromacs 3.3.3 that you get with gromacs 4.x. Hence it seems that the VACF in gmx 4 is correct. It also makes more sense intuitively.

#3 Updated by Vasudevan Venkateshwaran about 9 years ago

Hi, Yeah I managed to get the correct curve with gromacs 3. Thanks a lot.

#4 Updated by Rossen Apostolov about 9 years ago

So, can this be closed?

#5 Updated by Erik Lindahl about 9 years ago

Since the VAC appears to be correct now, and the reporter haven't said anything else I'm closing this bug - this might have been a build error in the submitters version 3.3.3.

Also available in: Atom PDF