[gmx-users] Regarding RMSD graph nalysis

[Justin Lemkul]

On 9/17/12 7:02 AM, naga sundar wrote:
> Dear gromacs users
>
> I run  20 ns MD simulation for protein mutant complexes. In RMSD analysis
> at ~9 ns sudden increase in the deviation was observed from 0.2 nm to 1.7
> nm and immediate fall was observed.
>
> I rerun the 20 ns simulation for the same molecule with same procedure.
> While analyzing the RMSD sudden increase in the deviation was observed (0.2
> nm to 1.7 nm) at the simulation period of ~12 ns.
>
> So i want to know  the reason
>
> For sudden rise and fall in RMSD values.
>
> Next, in first MD run this deviation was observed at ~9 ns (0.2 nm to 1.7
> nm) but while rerun the molecule with same procedure this deviation was
> observed at ~12 ns.
>

It sounds to me like your trajectories haven't correctly accounted for 
periodicity.  g_rms does not elegantly handle sudden jumps across the box.  You 
will need to work with trjconv.  Usually a simple trjconv -pbc mol -ur compact 
-center does the trick for a simple protein, but for complexes it is often more 
difficult.

http://www.gromacs.org/Documentation/Terminology/Periodic_Boundary_Conditions#Suggested_trjconv_workflow

-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

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