[gmx-users] trjconv and g_filter

[Justin A. Lemkul]

Yulian Gavrilov wrote:
> 
> Dear Gromacs users,
> 
> What type of conversion of coordinates is better to use?
> 
> I have got 100 ns MD simulations for unmodified ubc7 protein and 
> ubiquitinated ubc7 protein and converted it's trajectories by using 
> trjconv and g_filter:
> 
>          1.*trjconv* -s md100ns.tpr -f traj.xtc -o traj_noPBC_nojump.xtc 
> -pbc nojump -ur compact -dt 100
> 
>          2. *g_filter* -f traj.xtc -s md100ns.tpr -oh highpass.xtc 
> -nojump -b 25000 -dt 100 -fit -n calpha.ndx
> 
> I tried RMSF analysis for both variants.
> 
> In *trjconv* there are strange jumps (3, 4.5 angstroms) of residues 21, 
> 45, 100, etc. Most of these residues are within loops. It seems normal, 
> that they move more than others. But 4.5 angstroms – is it ok?
> 

Results are system-dependent.  I doubt anyone on the list knows what your 
outcome should be.  I see nothing wrong with a 4.5-nm RMSF; I've seen higher 
values for particularly flexible regions of different proteins.

> As I understand *g_filter* is used more to make good movies, but without 
> it I get strange RMSF.
> 
> Can I use g_filter instead of using trajconv or after trajconv?
> 
> 

My gut feeling is no.  g_filter modifies the coordinates to give a smooth visual 
representation; these are not the actual coordinates produced by the MD 
simulation.  Just because it gives some aesthetically pleasing result does not 
mean that result is right :)

-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

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