drift in protein membrane simulation. membrane and protein one way, water the other way

From: maria goranovic (mariagoranovic_at_gmail.com)
Date: Mon Aug 20 2007 - 05:20:06 CDT

Hi,

I have a solvated TM protein membrane simulation system of ~ 80000 atoms.
After about 40 ns of simulation, the protein-membrane complex drifts in the
-z direction (along the bilayer normal) by about 7-8 Angstroms. The bulk
solvent, however, drifts in the opposite direction by approximately the same
distance. The overall center of mass does not drift. Please see the attached
graph for the center of mass progression of the different components,

>From my naive knowledge, systematic drift is common in simulations, and can
be corrected by recentering and reorienting the trajectory after the
production runs. But is it possible to have drift of different components of
a simulation in different directions ? Is something wrong in the simulation
?

I have used SHAKE with a timestep of 2fs (SHAKE was applied to all hydrogen
atoms). Electrostatics and vdw energies are calculated every step. All the
other parameters are typical. The config file is at the end of the email.

Thank you for the help,

-Maria

-- 
Maria G.
Technical University of Denmark
Copenhagen
-----------------------------------------------------------
# NAMDConfiguration file
set     pmex    125
set     pmey    125
set     pmez    100
##  .. read structure and parameters here
set temp 310
outputEnergies  5000
outputTiming    5000
outputPressure  5000
xstFreq         5000
dcdFreq         5000
wrapAll         on
rigidbonds     all
timestep        2
nonBondedFreq   1
fullElectFrequency 1
stepsPerCycle   20
exclude         scaled1-4
1-4scaling      1.0
switching       on
switchDist      10
cutoff          12
pairlistdist    14
Pme             on
PmeGridsizeX    $pmex
PmeGridsizeY    $pmey
PmeGridsizeZ    $pmez
#########################################
# PRESSURE AND TEMPERATURE CONTRON
#########################################
# temperature
langevin                on
langevinDamping         5
langevinTemp            $temp
langevinHydrogen        no
useflexiblecell     yes
langevinPiston          on
langevinPistonTarget    1.01325
langevinPistonPeriod    100
langevinPistonDecay     50
langevinPistonTemp      $temp
binaryoutput    on
outputname system-1
run 100000

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:46:49 CST