/JSCB

Primary LanguageFortran

JSCB

This is a fork of an implementation of the JSCB algorithm developed by Azad and Lawrence (Nucleic Acids Research 2007, 35, 4629-4639) and further validated by Mehul et al. (Open Biology 2017, 7, 170094). Please cite these papers if you use this code. We have added a wrapper script that can handle draft quality genomes (i.e. with multiple contigs).

Installation

First download the repository using git clone, then add the repository's directory to your $PATH variable. The jscb file is a Fortran executable compiled for Linux x86. If you have some other system, you will have to recompile the jscb.f file. On Ubuntu or other Debian-derived linux distributions you will have to install a few modules before running:

sudo apt-get install build-essential libgfortran3

Usage

First annotate your genome (we recommend the Prokka pipeline). Then feed the resulting genbank file into the run_jscb.py script:

run_jscb.py --genbank input_genome.gbk --output_dir directory/path

In the output directory, there will be a tab-delimited table called genomic_islands_summary.tsv, which has the following columns:

  • Genomic island ID (GI_ID)
  • Contig
  • Start coordinate
  • End coordinate
  • Genes (comma-delimited list)

There will also be a file called hgt_genes_summary.tsv, which has the following columns:

  • Locus tag
  • Cluster ID
  • Cluster size (number of genes)
  • Gene length

JSCB makes clusters of genes based on nucleotide composition, and so the cluster with the most genes (probably cluster 1) likely represents genes that have not been horizontally transferred. There can be another cluster with, for example, translation factors and ribosomal proteins, that represent highly-expressed genes. Other clusters may be horizontally-transferred.

Original README.md contents below

Usage:

python jscb.py inputfile.gb

Input: JS-CB takes genbank(full) as input file

Output: JS-CB outputs three tab separated files:

  1. JSCB_output.tsv: Has three columns- Genomic Island (GI)-ID, GI-start co-ordinate and GI-end co-ordinate.
  2. JSCB_ouput.gi: Contains four columns- Gene number, Cluster id, Cluster size and Gene length.
  3. JSCB_output.clus: Contains four columns- Cluster id, Cluster size, Gene number, and Gene length.

Genomic island prediction can be found in JSCB_output.tsv file. JSCB_output.gi/.clus files contains cluster configuration information.

Requirements: The program requires Biopython. The binary JSCB file is compiled using F95 (Fortran95) compiler