LeapFrog
========

A set of tools that allows the genomic localization of (flanking
regions of) repetitive elements based on read-pair information.

analysis steps & data sets are:

* (1) input fastq
  The paired-end fastq files generated from the organism that you are
  interested in. This MUST be paired end!
* (2) reference fasta
  Contains the genome reference sequence
* (3) element_database
  The database is a multi-fasta file elements that are to be
  located. The software expects the sequence headers to have the
  following format: `>NAME#FAMILY`
* (4) bowtie2db for the reference fasta
  bowtie2 database based on (2)
* (5) bowtie2db for the element database  
  bowtie2 database based on (3)
* (6) get danglers
  the leapfrog script `lf_danglers` will run bowtie2 in the background
  and output a properly renamed fastq file containing the
  "danglers". A "dangler" is a read that does not map to the element
  database (3), but it's paired end mate does!
                                ____
                               /    \
                          =====      =====   <- dangler
      +========================+        
      | A sequence from the    |   
      | element database (3/5) |
      +========================+
   
  Check how to run the script using the `-h` parameter. This script
  takes as input the element bowtie2 database (5) and the input fastq
  (1).

* (7) map the danglers to the reference genome
  run a regular bowtie2 job mapping the dangler sequences (6) against
  the reference genome (2/4)
* (8) extract PFR's from the BAM alignment from (7)
  using the script lf_regionify. This script needs to be executed for
  each genome/sample separately. The output is a GFF file identifygin
  each PFR separately. The script splits PFR's based on family and
  orientation and tries to unmerge peaks that are close together. A
  score is assigned to each PFR.
* (9) compare PFR's between genomes
  This script (lf_findiff) is still very experimental. It takes a
  number of input GFF PFR files and determines which one overlap,
  followed by absence presence information.