[gmx-users] MD - analysis

[Justin A. Lemkul]

Kavyashree M wrote:
> Dear users,
> 
>     While analyzing an MD simulation run for 100ns, for temperature,
> pressure, volume and density, I got an output like this -
> 
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Temperature                     300      9e-05       -nan -0.000501989  (K)
> 
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                    0.99914      0.027       -nan  0.0174391  (bar)
> 
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Volume                      517.755     0.0094       -nan   0.010871  (nm^3)
> 
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Density                     1012.53      0.018       -nan -0.0216399  
> (kg/m^3)
> 
> Why is "nan" being shown for RMSD?

It means "not a number."  I've never seen that before from g_energy, but based 
on the extremely small error estimates, especially for the temperature, I'd 
suspect you're using extremely tight coupling which may be causing some error.

> when I checked the rmsd plot for the backbone, it had equilibrated 6ns,
> later on there were fluctuations.
> 

The RMSD presented by g_energy is the standard deviation of the data analyzed. 
It is unrelated to the RMSD of the protein.

> When I observed the trajectory in ngmx, after some time the protein molecule
> (the system consistes of 1 intact protein chain in water with few ions 
> to neutralize
> the charge) appeared fragmented. when I used the command
> "trjconv -s protein.tpr -f traj.xtc -o traj_atom.xtc  -pbc nojump"
> the entire molecule was intact throughout the simulation but was going 
> out of the
> box.
> 
> I had used dodecahedron cell for simulation this option was not present 
> in gnmx-
> box display options. Does the type of box selected in ngmx has effect on 
> whether
> molecule is inside the box or outside?
> 

In a periodic system, there is no such thing as "inside" and "outside" since the 
system is infinite.

> What I need to do to keep the molecule in the box and completely inside 
> the box?
> (I tried with -pbc options atom, res, mol - protein was in fragments in 
> all of them)
> 

trjconv -pbc mol -ur compact (and -center if you like).

> How can I visualize a dodecahedron box in ngmx?
> 

Never used ngmx, but with the trjconv command I've stated everything should show 
up quite nicely in VMD :)

-Justin

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

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

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