[gmx-users] Re: Error during CNP simulation

[Justin Lemkul]

On 10/30/12 3:28 AM, Bharath K. Srikanth wrote:
> Hi Chris
>
> My EM works just fine if I use a cubical box of water, so I'm baffled as
> to what's going wrong in this case.
>
> I tried the entire process again, this time removing position restraints.
> Let me describe my entire procedure on this occasion, once again:
>
> 1) I first assembled a bilayer of 128 DOPC lipids, in a box of dimension
> 7.5 x 7.5 x 7.5 nm3.
>
> 2) I then removed the water molecules, resized the box length by changing
> the bottom line of the coordinate file, and inserted a CNP to an
> appropriate location, about 7 nm above the bilayer.
>
> 3) Then, I performed an energy minimization, and the CNP did not move
> about. So far, so good.
>
> 4) I solvated the box with water using genbox, and then attempted an
> energy minimization (using steepest descents). This is where the problem
> arises.
>
> These are the coordinates from the input coordinate file, before the
> energy minimization:
>    129F16    C01 1793  10.600   3.474   3.330
>    129F16    C02 1794  10.438   3.758   3.294
>    129F16    C03 1795  10.381   3.588   3.947
>    129F16    C04 1796  10.795   3.703   3.387
>    129F16    C05 1797  10.564   3.308   3.584
>    129F16    C06 1798  10.289   3.506   3.415
>    129F16    C07 1799  10.322   3.939   3.528
>    129F16    C08 1800  10.655   3.430   3.872
>    129F16    C09 1801  10.840   3.479   3.620
>    129F16    C10 1802  10.641   3.958   3.453
>    129F16    C11 1803  10.342   3.876   3.850
>    129F16    C12 1804  10.610   3.991   3.756
>    129F16    C13 1805  10.675   3.736   3.950
>    129F16    C14 1806  10.307   3.397   3.721
>    129F16    C15 1807  10.862   3.787   3.699
>    129F16    C16 1808  10.172   3.686   3.649
>
> And these are the coordinates from the output coordinate file, within just
> TEN steps of energy minimization:
>   4849F16    C01 6513   1.015   0.049   5.350
>   4849F16    C02 6514   1.294  -3.135   4.063
>   4849F16    C03 6515   0.663  -1.872   3.948
>   4849F16    C04 6516   1.521  -1.365   3.815
>   4849F16    C05 6517   2.111  -0.464   4.030
>   4849F16    C06 6518   0.546  -2.205   6.177
>   4849F16    C07 6519  -0.418  -2.289   6.608
>   4849F16    C08 6520  -0.301  -1.054   4.991
>   4849F16    C09 6521  -0.154  -1.757   5.881
>   4849F16    C10 6522   0.595  -0.223   5.592
>   4849F16    C11 6523   1.450  -0.980   7.148
>   4849F16    C12 6524   0.055  -1.027   5.487
>   4849F16    C13 6525   0.020  -0.473   5.753
>   4849F16    C14 6526   1.492  -0.255   4.474
>   4849F16    C15 6527   0.926  -0.762   4.489
>   4849F16    C16 6528   0.147  -2.022   3.748
>
> I have no idea why this is happening- that the molecule could move so much
> seems ridiculous. Besides the em.mdp file, is something wrong with, for
> example, my method of resizing the box?
>

Do not manually resize the box.  Use editconf -box and -center as appropriate to 
place all your components within the unit cell.  While it may appear from 
visualization that your molecules are where you want them to be, they are not.

-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

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