Can i drop a single frame and continue?

Hey Guys,

I do have a question.
I simulated a molecules for 100ns and when I wanted to do RMSD and Gyration calculations from it, I got this error
“Reading frame 32000 time 63926.000 Floating point exception (core dumped)”

So if I am not mistaken this frame there is corrupt and therefore leads to problems.
Now, I am asking myself whether it is possible to just “skip” the frame somehow?

While preparing the no PBC-file i used trjcov and -dump 63926 but that obviously just removed every frame after that one :/

Is there a skip option or does anybody has an idea what I can do, besides simulating a new dataset?


Hi Max,

In the past we used a combination of gmx trjcat and gmx_rescue64 to reassemble a trajectory with bad frames excised. A long time ago one of my students wrote a simple script to semi-automate the process and put it up at GitHub - Becksteinlab/gmx_first_aid: A command line utility intended to assist in recovering data from corrupted gromacs trajectories. — sorry, there’s no documentation for the script beyond reading the script itself (and the script also needs to be adapted) but you might get an idea how to go about it. The source code for gmx_rescue64 is also in this repo.

I can’t help much more than pointing you to the repo. I hope you can make it work.


Dear Oliver,

thank you very much for the suggestions!

I will check our gmx_rescue and the repo soon!

did you solve it please ?