# Restarting simulation without the checkpoint file ( \*.trr doesn't contain coordinates) **URL:** \ **Category:** User discussions\ **Tags:** mdp-parameters, mdrun\ **Created:** [May 28, 2021, 5:52pm UTC](https://gromacs.bioexcel.eu/t/restarting-simulation-without-the-checkpoint-file-trr-doesnt-contain-coordinates/2243 "2021-05-28T17:52:59Z")\ **Posts on this page:** 2\ **Page:** 1
**Author:** ![hoda\_tb](https://avatars.discourse-cdn.com/v4/letter/h/9d8465/32.png) [@hoda\_tb](https://gromacs.bioexcel.eu/u/hoda_tb)\ **Post date:** [May 28, 2021, 5:52pm UTC](https://gromacs.bioexcel.eu/t/restarting-simulation-without-the-checkpoint-file-trr-doesnt-contain-coordinates/2243/1 "2021-05-28T17:52:59Z")
GROMACS version: 2019.6 GROMACS modification: No Hi everyone! I have a 200ns simulation and I want to continue a simulation from 160ns. I don’t have the checkpoint file because 160ns is not the last frame of the simulation. Since I need at least coordinates and velocities to continue the simulation, one way is to dump 160ns as a .gro file from the .trr file. Another way is to use grompp with .trr for -t. However, in my mdp file, coordinates are set to be written in xtc file and .trr file doesn’t contain coordinates. In this situation, none of the ways I said works! In the following lines, I shared my command lines and outputs. When I use -dump : **gmx\_mpi trjconv -f traj\_160ns.trr -s tpr.tpr -dump 160000 -o 160ns.gro** ( I extracted .trr of 160ns, named traj\_160ns.trr which contains velocities of 160ns, not coordinates) I get this output: _Select group for output_ _._ _._ _._ _Select a group: 0_ _Selected 0: ‘System’_ _trr version: GMX\_trn\_file (single precision)_ \*Last frame -1 time 0.000 \* ## _WARNING no output, last frame read at t=160000_ So there is no output! Or when I use grompp: **gmx\_mpi grompp -f prod.mdp -c conf\_160ns.gro -r conf\_160ns.gro -t traj\_160ns.trr -p topol.top -o test.tpr** I get this output: . . . _Note that mdrun will redetermine rlist based on the actual pair-list setup_ _Reading Coordinates, Velocities and Box size from old trajectory_ _Will read whole trajectory_ _trr version: GMX\_trn\_file (single precision)_ \*Last frame -1 time 0.000 \* _WARNING: Did not find a frame with velocities in file traj\_160ns.trr,_ - ``` all velocities will be set to zero!* ``` \*Last frame -1 time 0.000 \* _-------------------------------------------------------_ _Program: gmx grompp, version 2019.6_ _Source file: src/gromacs/gmxpreprocess/grompp.cpp (line 709)_ _Fatal error:_ _Did not find a frame with coordinates in file traj\_160ns.trr_ * * * It doesn’t find velocities and also needs coordinates which are not provided in the .trr file. Here is the check output of traj\_160ns.trr: _Command line:_ - gmx\_mpi check -f traj\_160ns.trr\* _Checking file traj\_160ns.trr_ _trr version: GMX\_trn\_file (single precision)_ \*Reading frame 0 time 160000.000 \* _# Atoms 720084_ \*Last frame 0 time 160000.000 \* _Item #frames Timestep (ps)_ _Step 1_ _Time 1_ _Lambda 1_ _Coords 0_ _Velocities 1_ _Forces 1_ _Box 1_ * * * These commands work perfectly if .trr file contains coordinates as well. But now my .trr doesn’t have coordinates and I haven’t found any solution to continue my simulation from an arbitrary frame. Is there anything that I can do? Or am I using something wrong here? Many thanks, Hoda ---
**Author:** ![alevilla](https://dub1.discourse-cdn.com/flex017/user_avatar/gromacs.bioexcel.eu/alevilla/32/439_2.png) [@alevilla](https://gromacs.bioexcel.eu/u/alevilla)\ **Post date:** [June 11, 2021, 12:09pm UTC](https://gromacs.bioexcel.eu/t/restarting-simulation-without-the-checkpoint-file-trr-doesnt-contain-coordinates/2243/2 "2021-06-11T12:09:17Z")
Hi, to extract velocities from trr file you can try [https://manual.gromacs.org/current/onlinehelp/gmx-traj.html](https://manual.gromacs.org/current/onlinehelp/gmx-traj.html) or [https://manual.gromacs.org/current/onlinehelp/gmx-trajectory.html](https://manual.gromacs.org/current/onlinehelp/gmx-trajectory.html) \Alessandra