[gmx-users] Corrupt trajectories

[Amil Anderson]

I have recently finished a very long MD simulation that created a 87 
Gigabyte .trr trajectory file and a 23 Gigabyte .xtc file.  When 
using trjconv to sample these trajectories the programs fails on the 
148,760th frame with a fatal error: "Can not determine precision of 
trn file".  The fatal error occurs at the same frame when I am 
reading either the .trr or the .xtc file.  Is there a way to get to 
the remaining 85% of the file?

I have tried skipping the first 149,000 frames using the -b option 
but still get caught at frame 148,760

Thanks for any assistance that can be provided.

Amil Anderson
Associate Professor of Chemistry
Wittenberg University
Springfield, OH  45501