[gmx-users] re: building lipid bilayer

[chris.neale at utoronto.ca]

I appreciate that english is not everybody's first language. Do you  
know a strong english speaker who can proof-read your questions before  
you submit them? I can gather from your statement "all head groups are  
coming like as slop" that things are not working, but that is about  
all that I can understand.

As for what I can suggest, a paper that I previously mentioned has a  
procedure that I have followed with good results (Takaoka et al  
Biophys J V79
2000 3118-3138.)

For the rest of this email I assume that dopc_min.pdb contains a  
single DOPC molecule? If not then the rest of this email makes no  
sense. Please be more specific in your future posts.

If you want to go with genbox, first use trjconv to align the long  
axis of your DOPC molecule in the z axis. Next you will need to write  
your own script to convert dopc_min.pdb into  
dopc_min_2_lipids_from_opposite_leaflets.pdb (basically reflect every  
atom about an XY plane (Z=constant) such that the closest approach of  
the tails of the two molecules should probably be around 0.25 to 0.4nm  
apart in the Z dimension). Next step is to use genbox to create a  
bilayer. If you do this you will notice that there is way too much  
space between monomers in the X and Y dimensions. You can reduce this  
space by reducing the X and Y dimensions of the box in the input  
structure file and repeating the genbox step.

When you solvate with genbox -cs water there will be water inside the  
membrane. There are lots of posts already on this list about how to  
deal with that. Come to think of it there are probably lots of posts  
on this list about this exact question... a quick search for "building  
lipid bilayer" (without quotes... your exact title) returned 25 hits,  
one of which was:

http://www.gromacs.org/pipermail/gmx-users/2003-July/006229.html

Please: search.