[gmx-users] cutoffs/PME

Michael Patra patra at lorentz.leidenuniv.nl
Thu Sep 11 22:37:01 CEST 2003

> i wonder if someone can give me references where GROMOS and/or GROMACS ff
> were used and the influence of different cutoffs on a lipid
> bilayer properties (like area per lipid, Scd... ) were studied??
> my simulations show that switching from twin-range 
> cutoff to PME results
> in a decrease of surface area.  i would expect that
> area will increase. Or is it a ff dependent?

We actually did a study of DPPC bilayers using the lipid parameters available
from the Gromacs website, and compared the results of PME and using different
cutoffs. The lipid parameters are based on the Gromos parameters, and basically
everybody (probably also you) is using these parameters. The study was
published as

    	Biophys. J. 84, 3636-3645 (2003)
What basically happens is that for our DPPC system the truncation leads to
artifically ordering. Ordering always decreases the area per lipid, so it is no
surprise that the area per lipids in our simulations goes down when switching
from PME to truncation. I can see no principal reason why in other situations
and/or using other force fields, truncation should not be able to destroy some
ordering, thereby increasing the area per lipid. What you need to compute for
both of your simulations (PME and cutoff) is

1) the order parameter of the chains (g_order, you apparently already did this)

2) two-dimensional radial-distribution function of your lipid molecules
  (unfortunately g_rdf has no option for two-dimensional rdfs, so you will
  need to write your own routine)

Only once you have those two plots, you can understand what is going on, and
whether it makes sense.


Dr. Michael Patra
Biophysics and Statistical Mechanics Group
Laboratory of Computational Engineering
Helsinki University of Technology
P.O.Box 9203 (Tekniikantie 14)
FIN-02015 HUT


Phone: +358-9-451 5724
Fax:   +358-9-451 4830 

More information about the gromacs.org_gmx-users mailing list