[gmx-users] MDrun rerun issue

[dodia]

Dear gromacs users,

I am new user and trying to learn about the software by running simple 
water box simulations. I did a 15ns NVT simulation of around 80 water 
molecules (after 1.5ns+1.5ns of NVT & NPT equilibration) at 300K and the 
resultant data seemed reasonably accurate wrt density and average 
temperature. Now I tried to extract 50 frames from the trajectory and 
stored them in a .gro file (with positions+velocities) and tried to use 
mdrun -rerun to crosscheck energies, but in the resultant log file, the 
temperature varies around 7000K, potential energy becomes positive and 
other values go haywire. Can anyone help me with this issue? I used the 
following commands after production calculation:


echo 1 | gmx trjconv -f prdv.trr -s prdv.tpr -vel -dt 100 -box 1.32 1.32 
3.5 -o frame.gro
gmx mdrun -s prdv.tpr -rerun frame.gro -g test.log

prdv is the file suffix for production simulation. prdv.trr contains all 
the positions, velocities and forces at every 1.5ps

Best Regards,
Mayank Dodia