Two perl scripts are useful for parsing through the results file and produce much shorter files which can be examined manually. These are useful when doing multiple program runs with the -m option in Co-Bind. -------------------------------------------------------------------- 1. parse_results.pl usage: perl parse_results.pl Example: perl parse_results.pl sample.run > parse_out The perl script generates a summary for each run. An example summary format is given below: START OF RUN: 17 # CONSENSUS ACGATACTGA # INFORMATION 5.357 (base e) # CONSENSUS GCATGTCCGG # INFORMATION 4.424 (base e) # >> Best iter for two units combined: 287 # >> Maximum conbined energy for both units (MaxUU): 0.847768 # >> MaxU1: 6.49107 MaxU2: 6.26696 The summary shows: a) The number of the run, b) Consensus obtained from the perceptron weights for the two perceptrons (CONSENSUS) c) The information content of aligned sites for the two perceptrons (INFORMATION) d) Iteration which gives the maximum value of the objective function ("Best iter for two units combined:") e) Maximum value of the combined objective function (MaxUU) f) Maximum value of objective function for each unit (MaxU1 and Max U2). --------------------------------------------------------------------- 2. sort_results.pl usage: perl sort_results.pl Example: perl sort_results.pl parse_out 10 > sort_out This perl script sorts the results according to the maximum objective function value (MaxUU), and outputs the summary of a certain number top scoing runs. For summary format, see above.