Under Construction! (Spring 2016)


Installation


Get The Code

Either visit our download page or clone our git repo.

$ git clone https://github.com/robinmeyers/transloc_pipeline.git

Add directories to $PATH in ~/.profile or ~/.bash_profile

$ echo 'export PATH=~/transloc_pipeline/bin:~/transloc_pipeline/R:$PATH' >> ~/.bash_profile


Reference Genomes

The pipeline requires both a fasta file and a bowtie2 index of your reference genome. The pipeline will search for these elements in the locations specified by two environment variables: $GENOME_DB and $BOWTIE2_INDEXES. The pipeline will throw an error if it cannot find the bowtie2 index files at $BOWTIE2_INDEXES/<reference id> and the genome fasta at $GENOME_DB/<reference id>/<reference id>.fa.

First decide on your paths (e.g. ~/genomes or /usr/local/genomes for $GENOME_DB) and then add required environment variables to in ~/.profile or ~/.bash_profile

$ echo 'export GENOME_DB=~/genomes' >> ~/.bash_profile
$ echo 'export BOWTIE2_INDEXES=~/genomes/bowtie2_indexes' >> ~/.bash_profile
$ source ~/.bash_profile

The code below is one way to set up the pipeline to run against the hg19 reference genome. Running this will take some time to download the bowtie2 index and write the fasta file.

$ mkdir -p $BOWTIE2_INDEXES
$ cd $BOWTIE2_INDEXES
$ wget ftp://ftp.ccb.jhu.edu/pub/data/bowtie2_indexes/hg19.zip
$ unzip hg19.zip
$ mkdir -p $GENOME_DB/hg19
$ bowtie2-inspect hg19 > $GENOME_DB/hg19/hg19.fa


Software Dependencies

There are lots. Sorry about that.

Bowtie2

Install Bowtie2

If on OS X and using homebrew:

$ brew tap homebrew/science
$ brew install bowtie2

Samtools

Install Samtools

IMPT: The perl module that uses the samtools installation requires an older version of samtools. From the (README)[http://cpansearch.perl.org/src/LDS/Bio-SamTools-1.43/README], “This is a Perl interface to the SAMtools sequence alignment interface. It ONLY works on versions of Samtools up to 0.1.17. It does not work on version 1.0 or higher due to major changes in the library structure.” Also, on some machines you may have to edit the Makefile and add -fPIC to CFLAGS variable (CFLAGS = -g -Wall -O2 -fPIC) so that it plays nice with the Perl module.

Perl >= 5.16

On OS X, I have had good success with the ActivePerl distribution. Perlbrew is a good option if you are on a Linux machine without root access.

Bioperl

Install Bioperl

Open an interactive cpan shell:

$ cpan

Find the most recent version of bioperl:

cpan> d /bioperl/
Distribution    CJFIELDS/BioPerl-1.6.901.tar.gz
Distribution    CJFIELDS/BioPerl-1.6.923.tar.gz
Distribution    CJFIELDS/BioPerl-1.6.924.tar.gz

Now install:

cpan> install CJFIELDS/BioPerl-1.6.924.tar.gz

or

cpan> force install CJFIELDS/BioPerl-1.6.924.tar.gz

The Samtools module is installed separately. When prompted enter the path where you installed Samtools earlier.

cpan > install Bio::DB::Sam

Other modules (all of which can be installed using cpan install commands)

cpan > install Getopt::Long Text::CSV List::MoreUtils Data::GUID Interpolation IPC::System::Simple Storable Switch::Plain

I may have missed a few. Missing modules can be installed the same as above.

ea-utils

ea-utils is used during library pre-processing. Install to somewhere in your $PATH. In this example I use ~/local/bin.

$ wget https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/ea-utils/ea-utils.1.1.2-537.tar.gz
$ tar xzvf ea-utils.1.1.2-537.tar.gz
$ PREFIX=~/local/bin make

SeqPrep

SeqPrep is also used during library pre-processing. Make sure the executable ends up in your path as well.

$ git clone https://github.com/jstjohn/SeqPrep.git
$ cd SeqPrep
$ make
$ cp SeqPrep ~/local/bin/

Then add SeqPrep to your $PATH.

R >= 3.1.2

Download and install R

Load an R session:

$ R

Install required packages:

> install.packages(c("argparser", "magrittr", "readr", "stringr", "plyr", "dplyr"))
> source("https://bioconductor.org/biocLite.R")
> biocLite(c("GenomicRanges", "BSgenome"))



Running the Pipeline


Download Sample Data

Download .zip

or

$ git clone -b tutorial-data https://github.com/robinmeyers/transloc_pipeline.git tutorial_data
$ cd ./tutorial_data

Pre-processing libraries

The Alt Lab primarily uses in-line barcodes (sequenced by the MiSeq at the head of the forward read) and then deconvolutes pooled libraries using this program. This script also calls an external tool to trim Illumina adapter sequence. If using Illumina multi-plex barcoding strategy, this script will not be useful except for trimming adapters, which is still recommended.

Starting from pooled library fastq files

Deconvolutes and trims adapters

$ TranslocPreprocess.pl tutorial_metadata.txt preprocess/ --read1 pooled_R1.fq.gz --read2 pooled_R2.fq.gz

Starting from deconvoluted fastq files

Just trims adapters

$ cd ~/transloc_pipeline/data
$ TranslocPreprocess.pl tutorial_metadata.txt preprocess/ --indir ./

Running the pipeline

$ TranslocWrapper.pl tutorial_metadata.txt preprocess/ results/ --threads 2



Post-processing


Filtering

The TranslocPipeline main output is the *.tlx file. One master tlx file will be generated per library. It can be thought of as similar to a sam file for NGS libraries. It contains output for every read in the library, and will serve as the starting point for all down-stream analyses. However, for most analyses, it will need to be filtered into only the required reads for the specific analysis. There is a default filtering that the pipeline will generate automatically. For any other filtering regime, the user will have to re-filter the master tlx file.

Available filters

- unaligned: No OCS alignments.

- baitonly: Bait alignment is either the only alignent in the OCS or only followed by adapter alignment.

- uncut: Bait alignment runs greater than some number of bases past the cutsite.

- misprimed: Bait alignment runs fewer than some number of bases past the primer.

- freqcut: Restriction enzyme site within some number of bases from the junction. (Fairly depricated.)

- largegap: More than some number of bases between the bait and prey alignments.

- mapqual: OCS had a competing prey junction.

- breaksite: Prey alignment maps into non-endogenous breaksite cassette. (Fairly depricated.)

- sequential: Junction occurs downstream on read from first bait-prey junction.

Re-filtering a library

$ TranslocFilter.pl results/RF204_Alt055/RF204_Alt055.tlx results/RF204_Alt055_refiltered.tlx --filters ""

Ex. 1 Keep duplicate junctions

$ TranslocFilter.pl results/RF204_Alt055/RF204_Alt055.tlx results/RF204_Alt055_refiltered.tlx --filters ""

Ex. 2 Keep all un-translocated reads

$ TranslocFilter.pl results/RF204_Alt055/RF204_Alt055.tlx results/RF204_Alt055_refiltered.tlx --filters ""

Using a config file

Hotspot Detection

Two methods of detecting hotspots.

Using MACS2 to do translocaiton peak detection

Must have MACS2 installed.

$ tlx2BED-MACS.pl
$ macs2

Using scan statistics script to call peaks

$ TranslocHotspots.R

Circos Plots

And other viz.