Continuing umbrella sampling simulation

User discussions
1

GROMACS version:2018
GROMACS modification: No

Hi guys, I have been trying to continue some umbrella sampling simulations. Due to computational time, some of my simulations are aborted. I am trying to continue the simulation from the checkpoint file using the command: (all my files are named as nvt.xtc, nvt.tpr etc.)

gmx mdrun -v -cpi nvt.cpt -deffnm nvt

but I am receiving an error: saying “2 of the 5 output files are absent nvt_pullf.xvg and nvt_pullx.xvg, or modified or named differently”

However, this is not the case, both of these two files are present in the working directory and have not been renamed or modified.

I am a bit confused as I know that this is the standard way to continue or restart a simulation from cpt files, please let me know if anything else is required to be done

2

You can either explicitly name all the output files (.xtc, .trr, .xvg, .edr, etc.) instead of using -deffnm or else, you can use the -noappend option. The problem would be that you’ll get discontinuous trajectories, which should be analysed together later on.

3

Thank you Ali for the suggestion, I think both options you have mentioned are very much viable, thanks