[gmx-users] diffusivity computations

[Debashis Dutta]

Hi,
       I have trying to estimate the diffusivity of an alanine molecule 
in water with the following .mdp file (for md simualtions).

; Preprocessing
title               =  Alanine with Water
cpp                 =  /lib/cpp
include             =
define              =

; Run Control
integrator          =  md
tinit               =  0.000
dt                  =  0.002
nsteps              =  250000
init_step           =  0
comm_mode           =  None
nstcomm             =  1
comm_grps           =

; Langevin Dynamics

; Energy Miimization

; Shell Molecular Dynamics

; Output Control
nstxout             =  500
nstvout             =  500
nstfout             =  500
nstlog              =  500
nstenergy           =  500
nstxtcout           =  500
xtc_precision       =
energygrps          =  System

; Neighbor Searching
nstlist             =  10
ns_type             =  grid
pbc                 =  xyz
rlist               =  0.9

; Electrostatics and VdW
coulombtype         =  PME
rcoulomb_switch     =  0
rcoulomb            =  0.9
vdwtype             =  shift
rvdw_switch         =  0
rvdw                =  1.4
DispCorr            =  no
table-extension     =  1
fourierspacing      =  0.12
fourier_nx          =  0
fourier_ny          =  0
fourier_nz          =  0
pme_order           =  6
ewald_rtol          =  1e-5
ewald_geometry      =  3d
epsilon_surface     =  0  optimize_fft        =  yes

; Temperature Coupling
tcoupl              =  berendsen
tc-grps             =  System
tau_t               =  0.1
ref_t               =  300

; Pressure Coupling
Pcoupl              =  no
pcoupltype          =  isotropic
tau_p               =  0.5 compressibility     =  4.5e-5
ref_p               =  1.0

; Simulated Annealing

; Velocity Generation
gen_vel             =  yes
gen_temp            =  300 gen_seed            =  173529

; Bonds
constraints         =  none
constraint_algorithm = lincs
unconstrained_start =  no shake_tol           =  0.0001
lincs_order         =  4  lincs_iter          =  4
lincs_warnangle     =  30
morse               =  no

; Energy Group Exclusions

; NMR Refinement

; Free Energy Perturbation

; Non-equilibrium MD

; Electric Fields

; User Defined Thingies


When I plot the mean square displacement, however I do not see a linear 
variation with time (which is what I expected). Moreover, the estimated 
diffusivity (using the g_msd command) looks reasonable (1e-5 cm^2/s) 
for runs shorter than 0.1 ns. If I run it further, the diffusivity 
keeps increasing to values of 1e-3 cm^2/s (for 10 ns runs). Any 
comments or sugestions?


___________________________________________________________________________