VMD-L Mailing List

From: Anton Arkhipov (anton_at_ks.uiuc.edu)
Date: Wed Dec 24 2008 - 15:01:38 CST

Hi Mike,

There is no automated way to turn al-atom lipids into a SBCG model.
The reason is that the correspondence between the all-atom and SBCG
lipids is not very straightforward: for example, one SBCG two-bead
"molecule" represents 2.2 DOPC lipids on average. What we have been
always doing instead, is to create a SBCG membrane patch from scratch.
It's really easy. I'm sending you a bunch of simple scripts to do
this. We will probably add those scripts to the SBCG tools in the VMD
distribution. Also, we are preparing a tutorial about SBCG, which
covers SBCG models for both proteins and membrane. The tutorial is
being tested now, so I hope we'll add it to the list of existing VMD
and NAMD tutorials in a few months.

Using the scripts I'm sending, you can create a planar patch of SBCG
DOPC lipids (use generate_patch.tcl). Then, use the psfgen script
build-dopc.tcl, together with the topology file lipid-ion.top, to
build a PDB and PSF. You can also convert some fraction of DOPC to
DOPS using mutate-to-dops.tcl, and then run psfgen on this new system
using build-mixture.tcl. One can easily change the scripts and
topology file to model some other lipids (of course, some additional
parameterization would be appropriate in such a case).

Please note that the script generate_patch.tcl creates a uniform, flat
array of SBCG DOPC "molecules". Before simulations, you may want to
equilibrate the membrane patch, so that the "molecules" are not packed
in a lattice. Also, one could add a little bit of randomization to the
script, so that positions of the SBCG beads are not assigned on a
uniform lattice, but are distributed a little bit around such lattice
sites.

Hope this helps. Best,

Anton.

On 24 Dec 2008, at 11:59, Dolan, Michael (NIH/NIAID) [C] wrote:

> Hello,
>
> Is there an automated way to turn an all-atom representation of a
> DOPC bilayer into a SBCG model?
>
> I would like to simulate a DOPC bilayer with embedded protein using
> SBCG, but cannot find a straightforward method to do this. The idea
> is to perfrom a similar experiment to that doen in Anton's wonderful
> paper involving BAR domains (Biophys J (2008) 95(6):2806-21.) I can
> probably write a few scripts to do this more manually, but wanted to
> check here first.
>
> Any help is most appreciated,
>
> Mike Dolan
> NIH
>
>
>
>
>