With the help of gthsplit one can split GenomeThreader output files 
containing intermediate results into multiple sets according to different 
criteria.

gthsplit is called as follows:

gthsplit [options] intermediate_files

If no intermediate_file is given as input, stdin is used instead. This allows
to use gthsplit in a UNIX pipe.

gthsplit offers the following options:

-alignmentscore split according to the overall alignment score (scr).
-coverage       split according to coverage (cov).
-range          set the precentage range used to create the sets. 
                the sets are created counting from 100% downwards 
                (including the upper bound and excluding the lower bound)
                The default range is 5.

-v
-gzip
-bzip2

The following shell script gives an example on how to use gthsplit to split up
a file ``results'':

#!/bin/sh
# split up into 4 sets according to coverage

gthsplit -range 25 -coverage results

# produces the files results.cov0-25 results.cov25-50 results.cov50-75 and 
# results.cov75-100

# split each set into 5 subsets according to the overall alignment score

for suf in cov0-25 cov-25-50 cov50-75 cov75-100
do
  gthsplit -range 20 -score results${suf}
done

# we end up with the following files:

# results.cov0-25.scr0-20
# results.cov25-50.scr0-20
# results.cov50-75.scr0-20
# results.cov75-100.scr0-20
# results.cov0-25.scr20-40
# results.cov25-50.scr20-40
# results.cov50-75.scr20-40
# results.cov75-100.scr20-40
# results.cov0-25.scr40-60
# results.cov25-50.scr40-60
# results.cov50-75.scr40-60
# results.cov75-100.scr40-60
# results.cov0-25.scr60-80
# results.cov25-50.scr60-80
# results.cov50-75.scr60-80
# results.cov75-100.scr60-80
# results.cov0-25.scr80-100
# results.cov25-50.scr80-100
# results.cov50-75.scr80-100
# results.cov75-100.scr80-100
