Welcome to DeepRVAT’s documentation!

Rare variant association testing using deep learning and data-driven gene impairment scores.

_Coming soon:_ Overview of the DeepRVAT methodaster

How to use this documentation

First, see Installation.

Visit Quick start if you want to skip the detailed documentation and dive right in, or if you want to quickly check that the package installed correctly.

To run DeepRVAT on your data, first consult Modes of usage then visit the section relevant to your use case.

For all modes, you’ll want to consult Input data and configuration.

Note also that for all modes of usage other than association testing with precomputed gene impairment scores, you’ll need to preprocess your genotype data, followed by annotating your variants.

To train custom DeepRVAT models, rather than using precomputed gene impairment scores or our provided pretrained models, you’ll need to additionally run seed gene discovery. See also the Practical recommendations for training.

Finally, consult the relevant section for your use case here.

If running DeepRVAT on a cluster (recommended), some helpful tips are here.

Citation

If you use this package, please cite:

Clarke, Holtkamp et al., “Integration of Variant Annotations Using Deep Set Networks Boosts Rare Variant Association Testing.” Nature Genetics. https://www.nature.com/articles/s41588-024-01919-z

Contact

To report a bug or make a feature request, please create an issue on GitHub.

For general inquiries, please contact:

brian.clarke@dkfz.de

eva.holtkamp@cit.tum.de

Contents:

• Installation
• Quick start
  • Customize pipelines
  • Run the preprocessing pipeline on your VCF files
  • Annotate variants
  • Example DeepRVAT runs
• Modes of usage
  • Precomputed gene impairment scores
  • Pretrained models
  • Model training
• DeepRVAT preprocessing pipeline
  • Output
  • Setup environment
  • Configure preprocessing
  • Running the preprocess pipeline
• DeepRVAT annotation pipeline
  • Output
  • Input
  • Requirements
  • Configure the annotation pipeline
  • Running the annotation pipeline on example data
  • Running the pipeline on your own data
  • Configuring the annotation pipeline
  • Gene id file
  • References
• Seed gene discovery
  • Input data
  • Configuration file
  • Running the seed gene discovery pipeline
• DeepRVAT configuration and input data
  • Configuration file
  • Input data formats
• Association testing with precomputed gene impairment scores
• Association testing using pretrained DeepRVAT models
  • Configuration and input files
  • Executing the pipeline
  • Running the association testing pipeline with REGENIE
• Training and association testing
  • Configuration and input files
  • Executing the pipeline
  • Using cross validation
  • Running the association testing pipeline with REGENIE
• Output file formats
• Cluster execution
  • Pipeline resource requirements
  • Cluster execution
  • Execution on GPU vs. CPU
• Practical recommendations for training
  • Model architecture
  • Training traits and seed genes
  • Variant selection
  • Variant annotations
  • Model ensembling
• Applying DeepRVAT to UK Biobank data
• API Reference
  • deeprvat

Indices and tables

• Index

• Module Index

• Search Page