PROFASI  Version 1.5
List of all members
extract_props Class Reference

Calculating run-time properties from traj files. More...

#include <extract_props.hh>

Detailed Description

This program works with a traj file generated during a run to calculate values of various observables at the points where configurations were saved during the run. In other words, it can regenerate the rt file if the frequency of writing into that file is the same as the frequency of writing configuration snapshots. It is useful, if, for instance, we want to find the value of a new measurable quantity at all the points along a run-time history. The program simply reads the traj file, sequencially restores the population to each of the configurations saved during the run (within a range, if specified), and calculates any observables as specified.

PROFASI "traj" files are meta-data files containing information about how to find configuration snapshots at different MC cycles during a run. They don't contain the actual data on the state of the population during the run. The data is stored in binary format in a bunch of "conf..." files in the same directory as the traj file.

In its present form, this program has two intended uses.


extract_props [OPTIONS] traj_file

where, [OPTIONS] could be one or more of:

If you need a single PDB file at a certain point in a run, this program could be a sledge-hammer to kill a mosquito. Use the program extract_snapshot instead.

Although this program reads a settings file, it only cares about instructions about Observables in the settings file. To interpret the contents of the binary "conf..." files (dereferenced in the "traj" file), it uses the header section of that file or information written in a "" file in the same directory as the conf file. In other words, whatever you write in the settings file about number of chains and sequences has no effect on this program.


Re-generating run-time information with new Observables

Situation: you have a finished run of some system. While analysing the data you realize, you want to analyze the correlation of some new observables which you did not think of when you did the run. For instance, in a two chain run, you notice all minimum energy PDB files showing the chains either both folded or both unfolded. You want a scatter plot with the radius of gyration of each individual chain to see how they correlate, and perhaps to filter the scatter plot based on values of other observables. But radius of gyration of the individual chains was not in your settings file when you started the run. You can use extract_props to retrieve the missing information. The two lines you would need in the settings file for the two missing Rg observables are:

new_obs Rg rg0 of_chain 0
new_obs Rg rg1 of_chain 1

After adding these lines (or any other observable related lines), run one of the following:

$ extract_props -op tmp_ -rt n0/traj
$ extract_props -op tmp_ -rt -avg n0/traj

The first example above generates the run time history file ("rt") from the trajectory file n0/traj. The newly generated rt file will have an output_prefix "tmp_", i.e., it will be called "tmp_rt". A file called "tmp_rtkey" will have information about which columns contain the new observables.The entire range of MC sweeps available inside the traj file will be used. The second example creates an rt file as well as an averages file from the data described in the traj file. The averages file will be called "tmp_averages". To create histograms of the new observables, you will need to use the program prf_his1d (see Creating PROFASI style histograms with prf_his1d).

Generating a sequence of PDB snapshots from a run

$ extract_props -op tmp_ -pdb -sf -a 4589999 -z 4600999 --every 10 n0/traj

This extracts the structure of the population at cycles starting from 4589999 until 4600999 using every 10th available snapshot. The structures in pdb format are stored in a single file (because of "-sf") called "tmp_frames.pdb". Different snapshots in the PDB file will be given different MODEL tags.

See Also
extract_snapshot, Commands for setting up measurements/Observables

The documentation for this class was generated from the following files:

PROFASI: Protein Folding and Aggregation Simulator, Version 1.5
© (2005-2016) Anders Irbäck and Sandipan Mohanty
Documentation generated on Mon Jul 18 2016 using Doxygen version 1.8.2