Open kyleabeauchamp opened 10 years ago
That would indeed be nice. It could be a good exercise for me in writing Sphinx documentation, question is how soon I will get to it! :)
BTW I'm currently using the command line to run our T4 lysozyme simulations.
Can you resume a trajectory and append to the same DCD file?
It currently backs up the file. Are DCD files trivially appendable? My impression was that the OpenMM DCDReporter overwrites the file if it exists.
Also, it is possible to resume trajectories. I'm saving the restart information to a JSON format and then compressing it.
So it resumes, but creates a new DCD file?
Yeah, that's right. If the DCD file being written already exists, the existing file is backed up to filename_#.dcd
where #
is the smallest number corresponding to a file name that doesn't exist.
The restart file is backed up as well.
Side notes: On a Titan I am getting 55 ns/day with your provided system. In order to convert your PDB to a Gromacs .top file using "pdbgmx", I had to rename several hydrogens and residue names "HIS" -> "HID". OpenMM then loads the Gromacs .gro and .top file without any issues. In the future I will convert the my Gromacs .itp files to OpenMM .xml files, which means we won't need to load Gromacs coordinates and topologies anymore.
It might be cute to have an online HTML doc that nicely groups all the different command line options that are available. This could even be done in an automated way.