Copyright (C) 2015 Leibniz Institute for Natural  Product  Research  and
Infection Biology -- Hans-Knoell-Institute (HKI).

This program comes with ABSOLUTELY NO WARRANTY. This is  free  software,
and you are welcome to redistribute it under certain conditions. See the
file COPYING, which you should have received along  with  this  program,
for details.

      ###########################################################
      # CASSIS - (C)luster (ASS)ignment by (I)slands of (S)ites #
      ###########################################################

CASSIS is a tool to precisely predict  secondary  metabolite  (SM)  gene
clusters around a given anchor (or backbone) gene.  Genes  encoding  SMs
tend to be clustered. Gene clusters are small groups of normally  up  to
20 genes; tightly co-localized, co-regulated, and participating  in  the
same metabolic pathway.  The  predictions  are  based  on  transcription
factor binding sites  shared  by  promoter  sequences  of  the  putative
cluster genes.


usage: cassis.pl <parameters>


required parameters:

   --annotation, -a <file>

      Genome annotation file. A simple text file in tabular format  with
      at least five columns. These are:

      gene <string> | contig <string> |  start  position  <int>  |  stop
      position <int> | strand <+ or->.

      The column separator must be tab (\t). The annotation file must be
      sorted ascending by contig, start and stop. Start  positions  must
      be smaller than stop positions. Contig names have to coincide with
      the genome sequence file.

   --genome, -g <file>

      Genomic sequence  file.  A  multiFASTA  file  containing  the  DNA
      sequences of all contigs of the  species.  Contig  names  have  to
      coincide with the annotation file.


optional parameters:

   --anchor, -b <ID>

      Feature ID of the clusters anchor gene. The  ID  has  to  coincide
      with the annotation file. This gene will be the starting point  of
      the cluster prediction.

   --cluster, -c <name>

      Name of the gene cluster to predict. Creates a sub-directory  with
      the given name to separate predictions for different clusters  but
      same species (in the same working directory). Will be set  to  the
      cluster's anchor gene ID, if ommitted.

   --dir, -d <directory>

      Working directory. All directories and files generated  by  CASSIS
      will go in here. Choosing different directories for e.g. different
      species might be a good idea.

   --fimo, -f [<p-value cut-off>]

      Enables the motif search with FIMO.  Setting  a  maximum  p-value,
      which will restrict the number of binding sites found by FIMO,  is
      optional. If no p-value cut-off is specied, the default  value  of
      0.00006 will be used. To run the motiv search, the MEME suite must
      be    installed    on    your     system.     See     http://meme-
      suite.org/doc/download.html

   --frequency, -fq <frequency cut-off between 0 and 100>

      By default, CASSIS will skip the cluster prediction for a  certain
      motif, if this motif results in a number of binding sites found in
      more than 14 % of all promoters in the genome.  By  changing  this
      paramater you may set a different frequency cut-off.

   --gap-length, -gl <0-5>

      Sets the maximum number of promoters  in  a  row  WITHOUT  binding
      sites, which are allowed inside a cluster prediction. Allowed  are
      0, 1, 2, 3, 4, and 5. The default value is 2. Only applies if  the
      cluster prediction has been enabled (see parameter --prediction).

   --help, -h

      Show the help page, which you are currently reading.

   --meme, -m [<e-value cut-off>]

      Enables the motif prediction around the  anchor  gene  with  MEME.
      Setting a maximum e-value for found motifs is optional. MEME  will
      stop, if it reaches the e-value, regardless if any motif has  been
      found or not. If no e-value  cut-off  is  specified,  the  default
      value of 1.0e+005 will be used.  The  cut-off  -1  has  a  special
      meaning: This means NO cut-off and MEME will  not  stop  until  it
      finds at least one motif, regardless of its e-value.  To  run  the
      motiv prediction, the MEME suite must be installed on your system.
      See http://meme-suite.org/doc/download.html

   --mismatches, -mm <allowed mismatches>

      Setting a maximum number of allowed mismatches (0-3)  per  binding
      site sequence is optional. The default is 0. See --sitar.

   --num-cpus, -n <number>

      Set number of CPUs (or CPU cores) to  use.  The  motif  prediction
      (parameter --meme) and the motif search (parameter  --fimo)  steps
      will then run with multiple forks in parallel.  CASSIS  uses  only
      one CPU by default.

   --prediction, -p

      Enables the cluster prediction, including the motif prediction via
      paramter --meme and the motif  search  via  parameter  --fimo.  By
      default it's disabled.

   --sitar, -s <file>

      Alternative to MEME and FIMO: Enables the motif search with SiTaR.
      You have to provide a file in (multi) FASTA  format  with  binding
      site sequences of at least one transcription factor.

   --verbose, -v

      By default, only the most important information  is  printed.  Use
      this parameter if you prefer a more verbose output.


runtime (Intel Xeon, running at 2.7 GHz):

   Using 2 CPUs (via --num-cpus), predicting  the  cluster  to  a  given
   anchor gene takes about 40 min. Using 4 CPUs, it takes about 22  min.
   And using 60 CPUs, about 3 min.


command-line example:
cassis.pl --dir fungi/fumigatus/ --annotation fungi/fumigatus/A_fumigatus_Af293_version_s03-m04-r22_features.csv --genome fungi/fumigatus/A_fumigatus_Af293_version_s03-m04-r22_chromosomes.fasta --anchor Afu6g09660 --cluster Gliotoxin --meme 1.0e+005 --fimo 0.00006 -frequency 14 --prediction --gap-length 2 --num-cpus 2 --verbose


Contact: gianni.panagiotou@leibniz-hki.de, thomas.wolf@leibniz-hki.de

Cite: If you use CASSIS, please cite https://doi.org/10.1093/bioinformatics/btv713