onkopus 0.7.0

Variant interpretation framework to analyze and interpret genetic alterations in cancer

Onkopus: A modular variant interpretation framework

|

What is Onkopus?

Onkopus is an easy-to-use cancer variant interpretation framework to annotate, interpret and prioritize genetic alterations in cancer. Onkopus provides annotation for different mutation types including a wide range of features, including genomic, transcriptomic and protein information, biochemical features, pathogenicty prediction, functional effect prediction, and potential therapies.

Getting started

Install the Onkopus python package via pip:

python -m pip install onkopus

You can use Onkopus both as a command line tool and in python. All modules can also be installed locally as well.

Use Onkopus from the command line

Onkopus provides a command line tool to directly annotate variant files. Run the onkopus tool by specifying an input file (-i) and the genome version (-g) ('hg19'/'hg38'/'t2t'). Optionally pass an output file (-o) and specific modules (-m):

onkopus run -i somatic_mutations.vcf -g hg38 -o somatic_mutations.ann.vcf

(Note: This will annotate will all Onkopus modules, including potential therapies and drug classifications. Since this process may take a while, use it only for small to medium-sized variant files.)

To functionally annotate large VCF files, run

onkopus run -i somatic_mutations.vcf -g hg38 -m functional_annotation -o somatic_mutations.fn.ann.vcf

To annotate with specific Onkopus modules, select the module with the -m option:

onkopus run -i somatic_mutations.vcf -g hg38 -m alphamissense
onkopus run -i somatic_mutations.vcf -g hg38 -m revel,primateai

Liftover between reference genomes: Use the -g option to define the source file's genome version, and the -t option to define the target genome (here: hg19 to hg38).

onkopus run -i somatic_mutations.vcf -m liftover -g hg19 -t hg38

To test Onkopus with one of the built-in sample VCF files, run:

onkopus run -i somaticMutations.vcf -md test -g hg38 -o somaticMutations.ann.vcf

Liftover

To convert a variant file into another genome assembly, use the AdaGenes Liftover module integrated in Onkopus. Use g to specifiy the reference genome of the source file, and the -t option for the reference genome the file should be converted to (hg19/hg38/t2t).

onkopus run -i somaticMutations.vcf -m liftover -g hg19 -t hg38 -o somaticMutations.GRCh38.vcf

Install Onkopus locally

By default, Onkopus accesses the publicly available modules to annotate variants. It is also possible to install Onkopus completely locally, e.g. to process sensitive patient data.

To install Onkopus locally, make sure that you have installed Docker on your host system.

To change the configuration to use locally running modules, configure the following environment variables before you start Onkopus:

export ONKOPUS_MODULE_PROTOCOL=https
export ONKOPUS_MODULE_SERVER=localhost
export ONKOPUS_PORTS_ACTIVE=1

In the next step, you can install the Onkopus modules individually or all together. To install an Onkopus module, run onkopus install -m followed by the module identifier, e.g.

onkopus install -m clinvar

Run onkopus list-modules to get a complete list of all available modules. To install all available modules in one step, run

onkopus install -m all

(Note: This will install all available Onkopus modules. The required databases are installed as soon as you start Onkopus.)

After you have installed all the modules you need, you can start Onkopus with

onkopus start

To make sure that the installation and start of the modules was successful, you can run

docker ps

If the Onkopus modules have been successfully started, you should see an overview of the running modules.

To stop all Onkopus modules, run

onkopus stop

Use Onkopus from Python

You can also use Onkopus to directly load your variant data in your Python code as an AdaGenes biomarker frame. (Note: This will load the entire variant data into memory and is thus only applicable for small to medium sized variant files.) Use Onkopus from Python by running the full annotation pipeline or instantiate custom Onkopus clients and calling process_data:

Annotate variants with all modules

For a complete variant annotation, including functional and clinical annotation, run

import onkopus as op

bframe = op.read_file('./somatic_mutations.vcf', genome_version="hg38", input_format="vcf")
bframe = op.annotate(bframe)

Annotate with specific modules

import onkopus as op

genome_version="hg38"
bframe = op.read_file('./somatic_mutations.vcf', input_format='vcf')

# Annotate with ClinVar
client = op.ClinVarClient(genome_version=genome_version)
bframe.data = client.process_data(bframe.data)

# Annotate with AlphaMissense
client = op.AlphaMissenseClient(genome_version=genome_version)
bframe.data = client.process_data(bframe.data)

op.write_file('./somatic_mutations.annotated.vcf', bframe)

License

GPLv3

Documentation

A detailed documentation on how to use the Onkopus web application, the CLI and the python package can be found on the public Onkopus website.

Public version

A public instance of Onkopus Web is available at https://mtb.bioinf.med.uni-goettingen.de/onkopus.