[gmx-users] weird problem about editconf and genbox

Justin A. Lemkul jalemkul at vt.edu
Wed May 19 20:49:01 CEST 2010

Lin Xu wrote:
> Hi, I have two questions concerning generating a solvent box for my protein
> using gromacs 3.3.1.

Any particular reason you're using software that is over four years old? 
Gromacs 4.0.7 will give you a major speed upgrade, and lots of new features.

> 1) I used to use editconf to generate a dodecahedron box, then use genbox. It
> used to work fine: $editconf -f protein.gro -o protein_box.gro -d 0.9 -bt
> dodecahedron -c $genbox -cp protein_box.gro -cs spc216.gro -p example.top -o
> protein_box_water.gro
> But recently I found that using exactly the same command and input files, and
> the same program installed, the protein is not centered in the box (I
> download the structure and view it in pymol and vmd). I tried other box
> types, for triclinic and cubic, it seems -c flag can be recognized, but not
> for octahedron (nor dodecahedron). I wonder what things could be wrong.

Nothing is wrong.  Most visualization software defaults to a triclinic 
representation.  If you want the dodecahedral unit cell properly displayed you 
need to use trjconv -ur compact.

> 2) I have crystal structure for wild type protein, when I generated solvent
> box for it, genbox didn't put waters in active site. Then I made two
> single-point active site residue mutations using pymol: for a mutant
> Thr->Ala, no water is put in active site neither, but for mutant Thr->Ser, a
> water molecule is found in the active site. The presence of waters
> drastically changed my simulation results. This is again different from what
> I got in the past. I used to work with those three proteins and no waters
> were added in their active sites. I assumed mutant Thr->Ser shouldn't have
> water, since genbox delete extra waters according to VanderWaals radii, and
> the three proteins (one real structure and two models built up by pymol) are
> so highly similar. I wonder why the program decide to treat them differently,
> this time?

I don't know exactly what's going on, but the behavior of genbox can be 
influenced by the values of -seed and -vdwd.  Otherwise, if water should or 
should not be in the active site, sufficiently long restrained equilibration may 

> I also tried using different box types and box dimensions for mutant
> Thr->Ser, and found sometimes a water added, but sometime not. Is this
> inconsistency normal?

Different box types and dimensions will influence the way genbox places cubes of 
water in the active site.  It's not an inconsistency if you're manipulating 
other factors.


> Sorry for the long email. Those problems seem really weird to me. Any
> guidance or hint would be highly appreciated. Thank you very much,
> Lin Xu
> Ph.D. Candidate Boxer lab, Chemistry Department Stanford University Stanford
> CA 94305 (650)6448383


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list