[gmx-users] bilayers with genconf

[Luca Mollica]

> I think building bilayers with genconf works quite well.
> What we have done is align a lipid in a square box along the z-axis, say 
> 1x1 nm to avoid overlap between lipids, use genconf with the random 
> translation and random rotation option (rotate around z, translate along 
> z by rand(0.5 nm)), build a single leaflet of 64 lipids.
> Rotate (don't mirror) and translate that leaflet to make the other 
> leaflet (editconf).

OK, I've the confirmation that I've done everything right(this is 
exactly what I have done until now), but there was something not 
convincing me........

> The bilayer will be too large because of the 1x1 nm initial box, but you 
> can scale to the approximately right box size with editconf -scale.

..and this was that one!  :-(
The bilayer seemed to be rather big if related to some pdb membrane 
files downloaded, and now I know the problem......

> Minimizing restores the geometry, add water with genbox, remove water 
> with a script from the hydrocarbon interior, equilibrate.

that's the other part of the game!


Luca