[gmx-users] make_hole vs. genconf

[chris.neale at utoronto.ca]

> I'd like to build some lipids around a peptide.  Can I do so using  
> genconf?  (I know genconf can be used to build a membrane.  I'm  
> asking if genconf can be used to build a membrane around a solute,  
> like a transmembrane protein?)

"around"? No.

> I'm also aware of the make_hole tool.  My question here is:  Does  
> make_hole create a hole of specified dimensions and orientation,  
> according to the solute coordinates?  And can you specify exactly  
> where in the membrane you want the whole?

Make_hole doesn't actually make a hole. You need to make the  
premilinary hole yourself. The general problem is that if you take out  
all lipids that contact the protein, then you will have a huge vaccuum  
aroung the protein. So first you take out all the lipids whose  
headgroup clashed  (manually) then run make_hole to bend the tails out  
of the space that is to be occupied by the protein, then add the  
protien.

You will need to install gromacs-3.1.4 to do this.

> So basically what I'm asking is:  If you want to make a starting  
> structure (or several structures) with the solute (a peptide) in a  
> specific part of the membrane, is it better to set the desired  
> coordinates of the peptide first and then build a membrane around  
> it?  Or, is it wiser to start with an equilibrated membrane and use  
> make_hole to create your hole first at the desired coordinates, then  
> insert the peptide?

1. Build your membrane (or find coords somewhere)
2. equilibrate it
3. Use make_hole or inflategro (I have never tried this one)
http://moose.bio.ucalgary.ca/index.php?page=Translate_lipdis
4. re-equilibrate

Reason not to do it all at once is because you want to limit your  
protein's exposure to vaccuum, although I suppose you could use  
position restraints to solve this issue (heavy atom on the protein and  
a z-only on the lipid phosphate).

Chris