HiC-inspector

About

High-throughput conformation capture (Hi-C) allows to study the three- dimensional architecture of whole genomes through the detection of long range chromosomal interactions. We developed HiC-inspector, a bioinformatics pipeline to facilitate the analysis of HiC dataset. Our tools performs read alignment, filtering of the reads that are in the DNA fragment size window around the restriction enzyme sites, counting of interactions with a user- defined resolution, and generation of contact matrix and heatmaps with a complete mapping of the interaction in the reference genome. HiC-inspector is publicly available open source, and can be used for paired-end sequencing from different platforms.

Authors

HiC-inspector is written by Giancarlo Castellano, François Le Dily, Antoni Hermoso Pulido and Guglielmo Roma from Bioinformatics Core and Miguel Beato's Lab @ CRG.

License

This software is distributed under the terms of GPL 3.0

Source

You can grab the last code from:

https://github.com/HiC-inspector/HiC-inspector

Citation

Castellano, G., Le Dily, F., Hermoso Pulido, A., Beato, M., & Roma, G. (2015). Hi-Cpipe: a pipeline for high-throughput chromosome capture. bioRxiv. Cold Spring Harbor Labs Journals. doi:10.1101/020636

Contact

If you have any comments, suggestions, questions, bug reports, etc., you can submit an issue in Github or feel free to contact: toni.hermoso@crg.eu

PLEASE attach your command line and log messages if possible.

README

Introduction

With the improvement of sequencing techniques, chromatin digestion with a restriction enzyme, ligation of the resulting sticky ends followed by high throughput sequencing is getting popular to study genome-wide DNA-DNA interactions. We present a novel pipeline, named HiC-inspector, for identifying DNA interacting sites.

Requirements

• awk, zcat, GNUTools (they are commonly in most Linux distributions)
• perl. Required modules are normally present in most Linux distributions. Alternately, used modules are: ''Getopt::Long'', ''Data::Dumper'', ''Pod::Usage'', ''POSIX 'strftime''', ''Benchmark'' and ''Parallel::ForkManager''.
• R http://www.r-project.org (originally tested in 2.13 and 3.2) Install also the following R libraries:
  • gplots http://cran.r-project.org/web/packages/gplots/index.html
  • RColorBrewer http://cran.r-project.org/web/packages/RColorBrewer/index.html
  • MCMCpack http://mcmcpack.wustl.edu/installation.html
  • Heatplus http://www.bioconductor.org/packages/release/bioc/html/Heatplus.html
  • HiTC http://www.bioconductor.org/packages/release/bioc/html/HiTC.html
• Bowtie http://bowtie-bio.sourceforge.net (originally tested in 0.12.7)
• Bedtools http://code.google.com/p/bedtools (originally tested in 2.10.1)

Install

Please check the file 'INSTALL' in the distribution.

conf.pl

• In this file you can change some of the path and parameters used by the application.
• Don't forget the trailing slash at the end the dir's path.

BEGIN {
    package main;

    %conf = (
        # Path to local tools
                'rdir'                  => "/soft/general/R-2.13/bin/",
                'bowtiedir'     => "/data/projects/hic/bin/",
                'bedtoolsdir'   => "/data/projects/hic/bin/",
                # Debug while running
                'debug'                 =>1
        );
}

1;

Usage

Usage: perl
hic-inspector.pl [-n missmatches] [-m multiplemappings] [-cf chrsizefile] [-df
designfile] [-sf selectfile] [-dd datadir] [-rd restrictiondir] [-dfo
dataformat] [-pd projectdir] [-g genome] [-fs fragmentsize] [-b bin] [-s step]
[-t test] [-u utils] [-pr processors] [-h help]

Options:

-n, -missmatches= Bowtie option -n: Max mismatches in seed (can be 0-3, default: -n 2) 
-m, -multiplemappings= Bowtie option -m: Suppress all alignments if >  exist (def: no limit) 
-pd, -projectdir= Directory where to write all the results. This folder is created by the pipeline, if it does not exist 
-df, -designfile= Input file describing the experimental design (tab separated text: 1st-column is sample_name, 2nd-column is read1_file, 3rd-column is read2_file, 4th-column is restriction_enzyme_file) 
-dd, -datadir= Directory containing data to be analysed. These can be raw reads in qseq or fastq format(compressed or not), or mapped reads in BED format 
-rd, -restrictiondir= Directory containing restriction enzyme sites to be considered in the analysis. Should be provided in BED format 
-dfo, -dataformat= Format of sequencing reads. Valid options are: qseq (default), fastq, and bed 
-g, -genome= Indexed genome file for the reads alignment 
-sf, -selectfile= Input file with user-defined genomic regions of interest (BED format) 
-fs, -fragment_size= Maximum expected fragment size 
-cf, -chrsizefile= Input file with chromosome sizes (tab separated text: 1st-column is chr, 2nd-column is size) 
-b, -bin= Genomic windows, or "bins", to count for chromatin interactions. Several bins can be provided as a comma separated string (e.g. -b 100000,1000000) 
-s, -step= Analysis steps to be performed. More steps can be provided either as comma-separated list [e.g. 1,2,3] or as dash-separated range [e.g. 1-3] 
Available steps: 
	STEP 1: copying files to local directory; 
	STEP 2: converting qseq file to FASTQ format; 
	STEP 3: mapping reads to genome; 
	STEP 4: converting mapping output to BED format; 
	STEP 5: filtering reads by proximity to restriction sites; 
	STEP 6: filtering reads for regions of interest; 
	STEP 7: combining mate filtered outputs; 
	STEP 8: calculating distances distribution between mate pairs; 
	STEP 9: generating contact matrix; STEP 10: analyzing contact matrix; 
-t, -test Test mode. Prints out commands without executing them 
-u, -utils Specify the genome release to be used among those already provided by HiC-inspector (e.g. hg19) 
-pr,   -processors=               Number of processors to be used -for allowing parallelization (default: 2)
-help This documentation. 

Example result

We provide an example result at: http://biocore.crg.cat/software/HiC- inspector/

This uses hg19 processed with hindIII and a couple of reads from SRR027956.

• For convenience, we provide hg19 and hindIII processed files here (md5sum: e3509b155ded6e3b5eb98e996752ba05).

Design file

Named design.GM.hindIII.hg19

GM.hindIII SRR027956_1.fastq  SRR027956_2.fastq  hindIII.hg19.bed

Take care to keep tabs instead of simple spaces.

Chromosome files

In utils directory you need to put a:

• chromosome sizes file ( e.g. chromsizes.hg19 )
• a directory with pre-built Bowtie genome indexes

We used fetchChromSizes script to create the chrom.sizes file for the UCSC database you are working with (e.g. hg19)

• BED file with restriction enzyme processed genomes.

You can use oligoMatch (ref), among other software, to process your genome with a restriction enzyme

oligoMatch hindIII.fa hg19.2bit hindIII.hg19.bed

hindIII.fa file would look this way:

>hindIII
AAGCTT

Example execution command

This is a command with the minimal number of parameters:

perl hic-inspector.pl -df design.GM.hindIII.hg19 -dd inputreadsdir -pd output/myproject.hindIII.hg19 -dfo fastq -u hg19 -b 1000000,10000000

Docker

The easiest way to test and use this tool is by using Docker.

• Create your own custom directories: input, output and utils

  • mkdir -p /path/to/my/input
  • mkdir -p /path/to/my/output
  • mkdir -p /path/to/my/utils

• Download SRR027956 SRA example in input (you can also use a 3rd party Docker image)

  • cd /path/to/my/input; docker run --rm -v "$(pwd)":/data -w /data inutano/sra-toolkit fasterq-dump SRR027956

• Prepare genome indices, chromosome sizes, etc. in utils directory. This must match assembly code used. We will use hg19 file provided above.

  • cd /path/to/my/utils; mkdir hg19; cd hg19; wget http://biocore.crg.cat/software/HiC-inspector/extra/hg19.tar.bz2; tar jxf hg19.tar.bz2

• We prepare a design file somewhere. Here we place it in /path/to/my/input directory

  • cd /path/to/my/input; wget http://biocore.crg.cat/software/HiC-inspector/extra/design.GM.hindIII.hg19

• Let's run the program

  • First we prepare a container: docker run -d -v /path/to/my/input:/input -v /path/to/my/utils:/utils -v /path/to/my/output:/output --name myhic biocorecrg/hic-inspector tail -f /dev/null
  • Then we execute the command: docker exec myhic perl hic-inspector.pl -df /input/design.GM.hindIII.hg19 -dd /input -pd /output/hindIII.hg19 -dfo fastq -u hg19 -b 1000000,10000000

• Once you don't need to run it for a while, you can stop the container:

  • docker stop myhic

• Note: fetchChromSizes and oligoMatch are also provided in another Docker image for convenience.

  • docker run biocorecrg/hic-inspector-kentutils fetchChromSizes hg19 > chromsizes.hg19
  • docker run -v /path/of/files:/share biocorecrg/hic-inspector-kentutils cd /share; oligoMatch hindIII.fa hg19.2bit hindIII.hg19.bed
  • You may want to edit resulting file (e. g. removing what is not a chromosome in chromosome files) and place them in /utils/hg19 (or equivalent assembly name)