Hydra-Multi - an SV discovery tool that incorporates hundreds of samples

Overview

Generate a config file.

======= Hydra-Multi is a paired-end read structural variant discovery tool that is capable of integrating signals from hundreds of samples.

Installation

Below are the requirements and instructions for installation of Hydra-Multi.

Requirements

samtools
pysam
set ulimit -f 16384

The ulimit determines the number of open file handles on a system.
This number must be larger than 4*number of possible chromosome-chromosome combinations in the respective reference.
For the human reference (hg19 b37), 16384 is the recommended ulimit.

Installing

git clone https://github.com/arq5x/Hydra
cd Hydra
make 
chmod +x scripts/*
sudo cp scripts/* /usr/local/bin
sudo cp bin/* /usr/local/bin

Testing Hydra-Multi

chmod +x hydra-multi.sh
./hydra-multi.sh test

Running Hydra-Multi

A wrapper script (hydra-multi.sh) can be used to automatiically run Hydra-Multi or each step may be performed manually. Both the automatic and manual executions begin by creating a stub file.

0. Generate a stub file.

========================== Start with a simple config file "stub" such as the one below:

$ cat config.stub.txt
sample1 /full/path/to/file/sample1.pos.bam
sample2 /full/path/to/file/sample2.pos.bam
sample3 /full/path/to/file/sample3.pos.bam

Automatic Execution

hydra-multi.sh can then then be used to execute subsequent steps:

./hydra-multi.sh run config.stub.txt

To obtain a parameter list for using the run script:

$./hydra-multi.sh run -h

usage:   hydra-multi.sh run [options] <stub_file>

positional arguments:
    stub file
        the stub file to create the configuration file, example on https://github.com/arq5x/Hydra
options:
    -t INT  Number of threads to use. [Default: 2]
    -p INT  The punt parameter for assembly, the maximum read depth allowed. [Default: 10]
    -o STR  The stub for the output file names

Manual Execution

1. Generate a config file.

==========================

HydraMulti needs a configuration file documenting the sample/libraries and the paths to their respective BAM files that will be input to SV discovery process.

The make_hydra_config.py script will inspect the alignments in each sample's BAM file to automatically create a complete config file documenting the statistics of the fragment library:

python scripts/make_hydra_config.py -i config.stub.txt
sample1 /full/path/to/file/sample1.pos.bam  374.23  12  3
sample2 /full/path/to/file/sample2.pos.bam  398.19  20  3
sample3 /full/path/to/file/sample3.pos.bam  401.78  23  3

Just redirect the output to a new, complete config file and you should be ready to go:

python scripts/make_hydra_config.py -i config.stub.txt > config.hydra.txt

2. Extract discordant alignments.

================================= Once you have created a configuration file for Hydra-Multi, you need to run the extract_discordants.py script to, you guessed it, extract the discordant alignments from your BAM files into BEDPE format for HydaMulti.

For each inout BAM file listed in your configuration file, extract_discordants.py will create a BEDPE file of the discordant alignments in the the same directory. For example, it will create a sample1.pos.bam.bedpe file for the sample1.pos.bam input file listed in the config file:

python scripts/extract_discordants.py -c config.hydra.txt -d <sample_name>

3. Run HydraRouter

================================= This routes all of the alignments on with the same chromosome/orientation set to the same file for assembly.

$ hydra-router -config config.hydra.txt -routedList routed-files.txt

4. Assemble SV breakpoint clusters

================================== Assembly of each chromosome/orientation set.

$ sh scripts/assemble-routed-files.sh routed-files-test.txt config.hydra.txt 1

5. Combine the individual SV assembly files into a single file.

=============================================================== Combine all of the chromosome/orientation sets back into one file.

$ sh scripts/combine-assembled-files.sh /full/path/to/assembled/files/ all.assembled

6. Finalize the SV breakpoint predictions.

===============================================================

$ scripts/forceOneClusterPerPairMem.py -i all.assembled -o all.sv-calls

7. Determine presence of the SV breakpoint predictions in samples.

=======================================================================

$ scripts/frequency.py -f all.sv-calls.final -d all.sv-calls.detail > all.sv-calls.freq

8. Change footprint intervals into breakpoint intervals.