BIAS-2015

Bitscopic Interpreting ACMG 2015 Standards

This software is dual licensed. You may choose the AGPL license or contact us for a commercial license to fit your needs.

Setup

See doc/variant_interpretation.txt for generating the required data files. Once generated, update the paths in the file config/requiredPaths.json to reflect the appropriate paths on your machine.

Alternatively, our version of the required data files can be downloaded here BIAS-2015-data

Users will need the files for their respective genome build.

BIAS-2015 exclusively uses Python3 standard libraries, you should not need to install any other dependencies.

To run the pipeline on a VCF file, you will first need to annotate it with Illumina Connected Annotations (ICA)

ICA (formerly Nirvana) is a dotnet software package that is free to Download online here https://support.illumina.com/downloads/illumina-connected-annotations.html

BIAS-2015 currently supports hg19/GRCh37 and hg38/GRCh38 classifications, as such you must annotate with either the ICA hg19/GRCh37 or hg38/GRCh37 dataset.

mkdir data
dotnet IlluminaConnectedAnnotations/Downloader.dll --ga GRCh37 -o Data

We annotate our VCF with ICA/Nirvana in this manner

dotnet IlluminaConnectedAnnotations/Nirvana.dll --cache Data/Cache --sd Data/SupplementaryAnnotation/GRCh37 --ref Data/References/Homo_sapiens.GRCh37.Nirvana.dat --in test.vcf --o test_ica

If you are running DRAGEN >4.2, then you likely already have the output .json file in your results directory. You can run BIAS-2015 directly on this json file.

Running the pipeline

Included in the repository is a set of test data. If the code is installed and setup correctly, the following commands should complete in <10 seconds.

python bias-2015.py test/data/bias-2015_test_file.json config/requiredPaths.json test_output.tsv

If you used the same data source files as the manuscript, you can verify the same output is generated with this command

diff test_output.tsv test/data/bias-2015-expected_test_output.tsv

To include your own classifiers, please provide the optional --user_classifiers flag

python bias-2015.py test/data/bias-2015_test_file.json config/requiredPaths.json test_output.tsv --user_classifiers my_classifiers.tsv

Classifier files are the same format as BIAS-2015 output! If you have many variants need to be updated, or you wish to update your variant classifications with your own script, we recommend you run the pipeline first to make a classifier template then update the template. Once you have your classifiers ready in the template file, rerun the pipeline.

python bias-2015.py test/data/bias-2015_test_file.json config/requiredPaths.json test_output.tsv
mv test_output.tsv my_classifiers.tsv

Either manually or programatically update my_classifiers.tsv to include your own ACMG classifiers. Then re run

python bias-2015.py test/data/bias-2015_test_file.json config/requiredPaths.json test_output.tsv --user_classifiers my_classifiers.tsv

Unit tests

The unit tests are extremely helpful for developing or debugging specific portions of the code. They are not required for running the pipeline nor do they provide any insight into the pipelines performance or validity.

We use pytest as our unit test framework, users will need it installed and accessible to follow these testing commands.

To run the unit tests go to the test directory (some tests leverage local pathing)

cd test

Then use pytest

pytest

If the unit tests are not working, there is likely something extremly wrong and we recommend re downloading the repository.

Contribution guidelines

• All python code goes into the src directory.
• Every file in the src directory should have a paired file in the test directory.
• All python functions should have some degree of unit test coverage. This can be accomplished by either directly testing the function (preferred) or by testing a head function which calls other functions.
• Include publication citations directly in the code whenever possible.
• "To debug code you must be more clever than the person who wrote it, so don't be too clever when writing you own code".

Who do I talk to?

Chris Eisenhart [email protected] Joel Mewton [email protected]