[gmx-users] Re: Membrane Simulation

[Justin Lemkul]

On 4/24/13 4:49 AM, Giuseppe wrote:
> Dear Justin,
>
> there were for sure some errors in the command, but even having solved them,
> the protein still move during the optimization. What I noticed is that the
> coordinates of the residues seems to decrease to zero and than "restart"
> from a point out from the box. I mean, taking the first atom of the first
> residue, after one EM the coordinates were:
>
> 1MET      N    1   6.100  15.879   0.231
>
> after the second EM were
>
> 1MET      N    1   5.469  15.262   0.228
>
> after the third EM were
>
> 1MET      N    1   4.870  14.676   0.225
>
> and so on till the thirteenth EM in which the coordinates were
>
> 1MET      N    1   0.321  10.201   0.207
>
> after this EM, protein moves off the membrane and the coordinates of the
> first atom of the first residue become:
>
> 1MET      N    1  12.791   9.867   0.206
>
> there is obviously something wrong with the x axis, but what?
> The box vectors in this case were  6.23910   6.17970   6.91950 and the
> complete editconf command was "editconf -f protein.gro -o protein_newbox.gro
> -box 6.23910   6.17970   6.91950 -center 3.20920 2.72175 6.2 -rotate 145 -45
> 55".

The only way that this is possible is if your protein is much larger than your 
box.  Note that the y-coordinate of the N atom of Met1 begins at nearly 16 nm, 
which is more than twice the box vector in the y-dimension.  I suspect that your 
protein is ill-suited to being positioned within a box of these dimensions.

Note too that centering on z=6.2 almost assuredly places your protein outside 
the box if the z-dimension is 6.9195 nm.

If there are further problems, separate the rotation and placement steps.  I 
don't recall off-hand which operation is done first, but this can be a source of 
complication.  Rotate your configuration first, then position it in a second step.

> I also tried to use a larger box, which vectors were 12 12 12, and in this
> case the complete editconf command was "editconf -f protein.gro -o
> proetin.newbox.gro -box 12 12 12 -center 6 6 8.75 -rotate 145 -45 55" and to
> decrease the scaling factor from 4 to 2, but it was useless. I'm sorry to
> bother you so much, but I'm stucked with my project. Thank you so much.
>

A box size of 12 is still incompatible with the coordinates shown above.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================