[gmx-users] Converting trajectory into tpr file

[Justin Lemkul]

On 11/5/12 7:25 PM, rainy908 wrote:
> Fellow Gromacs users:
>
> Suppose I run a 1 ns equilibration on my protein system prior to molecular dynamics, and I am converting the equilibration trajectory to a TPR file.  Correct me if I'm wrong here - I'd like to confirm that I can generate a TPR file by using frames accessed in the ENTIRE trajectory, or just certain frame(s), i.e. last snapshot.
>

grompp does no such thing.  If given multiple frames, it uses the first frame it 
encounters.  grompp will use the last frame of a trajectory provided with -t, 
unless overridden by the -time option, in which case it uses the frame that most 
closely corresponds to that time value, per the printed documentation.

> 1. trjconv -f protein_equil.xtc -s protein_equil.tpr -o protein_b4md.gro
> 2. grompp -f md.mdp -c protein_b4md.gro -p topol.top -o protein_md.tpr
> 3. editconf -f protein_md.tpr -n index.ndx -o protein_md.pdb
>
> As compared to using the very last frame in the equilibrated trajectory:
>
> 1. trjconv -f protein_equil.xtc -s protein_equil.tpr -o protein_b4md_1ns.gro -b 1000 -e 1000
> 2. grompp -f md.mdp -c protein_b4md_1ns.gro -p topol.top -o protein_md_1ns.tpr
> 3. editconf -f protein_md_1ns.tpr -n index.ndx -o protein_md_1ns.pdb
>
> Does the former method output a structure that is an "ensemble average" based off of the equilibration time?  If it is indeed an average, should I use RMSD clustering to extract an actual structure accessed in the trajectory that is closest to the ensemble average?  Any suggestions would be greatly appreciated here.
>

There is no ensemble averaging in this protocol.  I'm not sure why you're trying 
to re-start a simulation from some arbitrary configuration during equilibration. 
  If you use some frame from within the equilibration, you'll have to start a 
new simulation from this point, which destroys the previous equilibration.

-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

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