Oviraptor detects oncoviruses and their integration sites in whole genome sequencing data

Usage

oviraptor tumor.bam -o results --host-fa hg38.fa

Installation

Via conda:

conda install -c vladsaveliev oviraptor

Installing conda is optional if you have the following tools installed and available in $PATH:

• python3
• minimap2
• samtools
• sambamba
• bcftools
• mosdepth
• snakemake

In this case you can install with:

git clone git@github.com:vladsaveliev/oviraptor.git
pip install oviraptor

Usage

The tool requires:

• a host whole genome seqeuncing BAM file as an input data (any human reference genome will work as the tool will extract reads from the file to realign),
• a human hg38 reference genome fasta file, which can be provided with --host-fa as follows:

wget --no-verbose https://hgdownload.soe.ucsc.edu/goldenPath/hg38/chromosomes/chr8.fa.gz
oviraptor test/tumor.bam -o test/results --host-fa test/chr8.fa.gz

The tool will also use a pre-packaged hg38 gene coordinates file to annotate the breakpoints. However you can override it with your own annotation file with --host-gtf, e.g.:

oviraptor test/tumor.bam -o test/results --host-fa test/hg38.fa --host-gtf Homo_sapiens.GRCh38.gtf.gz

If the --host-fa is not provided, the tool will attempt to download it from AWS-iGenomes using awscli into the output folder (results/reference, provided -o results), which might take a while and around 3G of space.

The tool can make use of multiple cores. To use 10 CPUs:

oviraptor test/tumor.bam -o test/results -t10

If you already know your candidate virus and just want to find the integration sites, use:

oviraptor test/tumor.bam -o test/results -v HPV18

If you don't need integration sites and just want to find viral content, use:

oviraptor test/tumor.bam -o test/results --only-detect

Development

To develop, install with -e flag to pip:

git clone --recursive git@github.com:vladsaveliev/oviraptor.git
pip install -e oviraptor

🦕