Gmx trjcat not splitting trajectories

User discussions
1

GROMACS version: 2021.5
GROMACS modification: No

Hello everyone,
I’m posting this so maybe somebody can help me with my issue.
I’m using a trajectory that is about 120 us long, representing various folded and unfolded states of a protein.
I used MDAnalysis and RMSD to discriminate between the folded and the unfolded states and generated two trajectories representing the two states.
My analysis implies that I divide the original 120 us long trajectory into four equal and smaller trajectories of 30 us each and I wanted to keep this identical division also for the folded/unfolded trajs.
With MDA I retrieved the intervals at which I should have cut the folded/unfolded trajectory using gmx trjcat. The issue is that the command functions only with the folded trajectory, it doesn’t run with the unfolded.
The command I used is

gmx trjcat -f unfolded-fulltraj_aligned_eq.xtc -o prova.xtc -b 0 -e 20000

that generates an empty file.
The total number of frames in the unfolded trajectory is 58119 and GROMACS output is

Reading frame 1 time 612700.000

Summary of files and start times used:

      File                Start time       Time step

unfolded-fulltraj_aligned_eq.xtc 612600.000 ps 100.000 ps

Reading frame 0 time 612600.000 Note that major changes are planned in future for trjcat, to improve usability and utility.
lasttime 0

Last frame written was -1, time -1.000000 ps

Do you have an idea of why I can’t split this trajectory? I tried loading it on VMD and it gave me no issues.

Thanks a lot!

2

Firstly, trjcat is used for concatenating several trajectories into one - not for splitting trajectories.

Secondly, according to your info the trajectory starts at t=612600 ps. With gmx trjcat -f unfolded-fulltraj_aligned_eq.xtc -o prova.xtc -b 0 -e 20000 you state that you should end the reading at 20 ns, which is before your trajectory has started.

I would suggest using trjconv instead. I’m not sure about the exact command, but something along the line of gmx trjconv -f fulltraj_aligned_eq.xtc -o prova.xtc -split 30000000 would hopefully work. However, I guess the splitting might not work perfectly with your starting time. Experiment and see.