merge_MD_batches.py

This function compiles the outputs of MD_batch_rin.py across multiple directories, so that entire trajectories can be broken up into sections and analysed “in parallel” for speed.

Note that the MD_batch_rin outputs in each directory must have the standard file names.

Help message 

usage: merge_MD_batches.py [-h] [-s SEED] [-dirs [DIRS ...]] [-addwat]

merge data from batch processed dirs

options:
  -h, --help           show this help message and exit
  -s SEED, -seed SEED  seed ch:ID[,ch:ID,ch:ID]
  -dirs [DIRS ...]     directories to merge data from
  -addwat              add waters to batch rin for each frame

Inputs

  • -s: seed, specified and used exactly the same way as in individual processing
  • -dirs: list of directories to combine MD outputs from
  • -type: which of probe and arpeggio to use to generate the RINs
  • -addwat: include the median number of waters in the batch_res_atoms files
    • off by default for now because distance analysis can be slow and so far we’ve mainly used this function for looking at the summary statistics

Outputs

  • batch_res_atoms.dat: res atoms file containing all non-water groups that have contacts in any of the given frames.
    • if -addwat used, a batch_res_atoms.[frame].dat file will be created for each frame with the median number of waters added as well
  • batch_stats.dat: summary statistics table of contact counts for each identified functional group
  • batch_rawdf.pkl: python pickle file containing pandas dataframe with all contact count info for every functional group in every frame. can be used for data analysis/plotting/etc
  • batch_atominfo.pkl: lists of selected atoms for each functional group used to create batch_res_atoms.dat
  • waters_per_frame.dat: list of how many water molecules have contacts in each of the given frames