vrs-python

VRS-Python provides Python language support and a reference implementation for the GA4GH Variation Representation Specification(VRS).

Information

Releases

Development

Features

• Pydantic implementation of GKS core models and VRS models
• Algorithm for generating consistent, globally unique identifiers for variation without a central authority
• Algorithm for performing fully justified allele normalization
• Translating from and to other variant formats
• Annotate VCFs with VRS
• Convert GA4GH objects between inlined and referenced forms

Known Issues

You are encouraged to browse issues. All known issues are listed there. Please report any issues you find.

Installing VRS-Python Locally

Prerequisites

• Python >= 3.10
  • Note: Python 3.12 is required for developers contributing to VRS-Python
• libpq
• postgresql

MacOS

You can use Homebrew to install the prerequisites. See the Homebrew documentation for how to install. Make sure Homebrew is up-to-date by running brew update.

brew install libpq
brew install python3
brew install postgresql@14

Ubuntu

sudo apt install gcc libpq-dev python3-dev

Installation Steps

1. Install VRS-Python with pip

VRS-Python is available on PyPI.

pip install 'ga4gh.vrs[extras]'

The [extras] argument tells pip to install packages to fulfill the dependencies of the ga4gh.vrs.extras package.

2. Install External Data Sources

The ga4gh.vrs.extras modules are not part of the VR spec per se. They are bundled with ga4gh.vrs for development and installation convenience. These modules depend directly and indirectly on external data sources of sequences, transcripts, and genome-transcript alignments.

First, you must install a local SeqRepo:

pip install seqrepo
export SEQREPO_VERSION=2024-02-20  # or newer if available -- check `seqrepo list-remote-instances`
sudo mkdir -p /usr/local/share/seqrepo
sudo chown $USER /usr/local/share/seqrepo
seqrepo pull -i $SEQREPO_VERSION
seqrepo update-latest

If you encounter a permission error similar to the one below:

PermissionError: [Error 13] Permission denied: '/usr/local/share/seqrepo/2024-02-20._fkuefgd' -> '/usr/local/share/seqrepo/2024-02-20'

Try moving data manually with sudo:

sudo mv /usr/local/share/seqrepo/$SEQREPO_VERSION.* /usr/local/share/seqrepo/$SEQREPO_VERSION

To make installation easy, we recommend using Docker to install the other Biocommons tools - SeqRepo REST and UTA. If you would like to use local instances of UTA, see UTA directly. We do provide some additional setup help here.

Next, run the following commands:

docker volume create --name=uta_vol
docker volume create --name=seqrepo_vol
docker-compose up

This should start three containers:

• seqrepo: downloads seqrepo into a docker volume and exits
• seqrepo-rest-service: a REST service on seqrepo (localhost:5000)
• uta: a database of transcripts and alignments (localhost:5432)

Check that the containers are running, by running:

$ docker ps
CONTAINER ID        IMAGE                                    //  NAMES
86e872ab0c69        biocommons/seqrepo-rest-service:latest   //  vrs-python_seqrepo-rest-service_1
a40576b8cf1f        biocommons/uta:uta_20210129b              //  vrs-python_uta_1

Depending on your network and host, the first run is likely to take 5-15 minutes in order to download and install data. Subsequent startups should be nearly instantaneous.

You can test UTA and seqrepo installations like so:

$ psql -XAt postgres://anonymous@localhost/uta -c 'select count(*) from uta_20210129b.transcript'
314227

It doesn't work

Here are some things to try.

• Bring up one service at a time. For example, if you haven't download seqrepo yet, you might see this:

  $ docker-compose up seqrepo-rest-service
  Starting vrs-python_seqrepo-rest-service_1 ... done
  Attaching to vrs-python_seqrepo-rest-service_1
  seqrepo-rest-service_1  | 2022-07-26 15:59:59 seqrepo_rest_service.__main__[1] INFO Using seqrepo_dir='/usr/local/share/seqrepo/2024-02-20' from command line
  ⋮
  seqrepo-rest-service_1  | OSError: Unable to open SeqRepo directory /usr/local/share/seqrepo/2024-02-20
  vrs-python_seqrepo-rest-service_1 exited with code 1

VRS-Python and VRS Version Correspondence

The ga4gh/vrs-python repo embeds the ga4gh/vrs repo as a git submodule for testing purposes. Each ga4gh.vrs package on PyPI embeds a particular version of VRS. The correspondences between the packages that are currently maintained may be summarized as:

vrs-python branch	vrs-python tag/version	vrs branch	vrs version
main (default branch)	2.x	2.x	2.x
1.x	0.8.x	1.x	1.x

⚠ Note: Only 2.x branch is being actively maintained. The 1.x branch will only be maintained for bug fixes.

⚠ Developers: See the development section below for recommendations for using submodules gracefully (and without causing problems for others!).

Previous VRS-Python and VRS Version Correspondence

The correspondences between the packages that are no longer maintained may be summarized as:

vrs-python branch	vrs-python tag/version	vrs branch	vrs version
0.9	0.9.x	metaschema-update	N/A
0.7	0.7.x	1.2	1.2.x
0.6	0.6.x	1.1	1.1.x

Developers

This section is intended for developers who contribute to VRS-Python.

Installing for development

Fork the repo at https://github.com/ga4gh/vrs-python/ and initialize a development environment.

git clone --recurse-submodules git@github.com:YOUR_GITHUB_ID/vrs-python.git
cd vrs-python
make devready
source venv/3.12/bin/activate

This setup includes pre-commit hooks. If you create a virtual environment manually, be sure to install the hooks yourself; otherwise, commits may fail during CI/CD checks:

source venv/3.12/bin/activate
pre-commit install

If you already cloned the repo, but forgot to include --recurse-submodules you can run:

git submodule update --init --recursive

Submodules

vrs-python embeds vrs as a submodule, only for testing purposes. When checking out vrs-python and switching branches, it is important to make sure that the submodule tracks vrs-python correctly. The recommended way to do this is git config --global submodule.recurse true. If you don't set submodule.recurse, developers and reviewers must be extremely careful to not accidentally upgrade or downgrade schemas with respect to vrs-python.

Alternatively, see misc/githooks/.

Testing

This package implements typical unit tests for ga4gh.core and ga4gh.vrs. This package also implements the compliance tests from vrs (vrs/validation) in the tests/validation/ directory.

To run tests:

make test

Running the Notebooks

The notebooks do not require you to setup SeqRepo or UTA from Install External Data Sources.

Running the Notebooks on Binder

Binder allows you to create custom computing environments that can be shared and used by many remote users.

You can access the notebooks on Binder here.

Running the Notebooks on the Terra platform

Terra is a cloud platform for biomedical research developed by the Broad Institute, Microsoft and Verily. The platform includes preconfigured environments that provide user-friendly access to various applications commonly used in bioinformatics, including Jupyter Notebooks.

We have created a public VRS-demo-notebooks workspace in Terra that contains the demo notebooks along with instructions for running them with minimal setup. To get started, see either the VRS-demo-notebooks workspace or the Terra.ipynb notebook in this repository.

Running the Notebooks with VS Code

VS Code is a code editor developed by Microsoft. It is lightweight, highly customizable, and supports a wide range of programming languages, with a robust extension system. You can download VS Code here.

1. Open VS Code.
2. Use Extensions view (Ctrl+Shift+X or ⌘+Shift+X) to install the Jupyter extension.
3. Navigate to your vrs-python project folder and open it in VS Code.
4. In a notebook, click Select Kernel at the top right. Select the option where the path is venv/3.12/bin/python3. See here for more information on managing Jupyter Kernels in VS Code.
5. After selecting the kernel you can now run the notebook.

Security Note (from the GA4GH Security Team)

A stand-alone security review has been performed on the specification itself. This implementation is offered as-is, and without any security guarantees. It will need an independent security review before it can be considered ready for use in security-critical applications. If you integrate this code into your application it is AT YOUR OWN RISK AND RESPONSIBILITY to arrange for a security audit.