[gmx-users] How do you properly use g_msd to calculate MSD of bilayer and account for monolayer motion?

[David Ackerman]

Hello,

I simulated a pure DPPC bilayer at 323 K under semi-isotropic conditions
and have a somewhat small area per lipid of .615 nm^2. I also wrote my own
code to calculate the average Mean Square Displacement and Diffusion
Coefficient of the lipids, taking into account the relative motions of the
monolayers, and am getting a result of around 5.4 x 10^-8 cm^2/s for a box
size of 256 lipids/leaflet and around 4 x 10^-8 cm^2/s for a box size of 64
lipids/leaflet. I am trying to compare my code directly to the built in
GROMACS result for g_msd but am unsure how to properly remove the monolayer
COM motions with g_msd. In the index file, I have a group of all the lipids
in one leaflet and then use g_msd with the flags -rmcomm (so that I can
remove the COM motion of a selected group) and -mol (so it calculates the
MSD for the COM of each lipid in the leaflet):

 g_msd -f md_0_1.xtc -s md_0_1.tpr -n diffusionIndex.ndx -o output -rmcomm
-b 10000 -mol

I then choose this one leaflet grouping for "Select a group to calculate
mean squared displacement for" and for "Now select a group for center of
mass removal", thinking that this would calculate the MSD of each lipid
with respect to the COM of the monolayer. It then says "Split group of 3200
atoms into 64 molecules", which seems correct, but then it freezes and does
not calculate. Is this calculation just extremely slow? If you used g_msd
for removing center of mass motion of the monolayers, could you please let
me know what groupings/flags you used to properly calculate the MSD? Many
publications also seem to have diffusion coefficients more along the lines
of 10x10^-8 for DPPC bilayers at 323 K, so does being off by a factor of 2
still seem reasonable? I appreciate any help you can provide.

Thanks,
David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20120619/7aa32bac/attachment.html>