Toolkit for benchmarking, merging, and annotating Structural Variants

📚 WIKI page has detailed user documentation.
🛠️ Developer Docs for the truvari API.
📈 See Updates on new versions.
📝 Read our Papers (#1, #2) to learn more.

💻 Installation

Truvari uses Python 3.6+ and can be installed with pip:

  python3 -m pip install Truvari 

For details and more installation options, see Installation on the wiki.

⏩ Quick Start

Each sub-command contains help documentation. Start with truvari -h to see available commands.

The current most common Truvari use case is for structural variation benchmarking:

  truvari bench -b base.vcf.gz -c comp.vcf.gz -f reference.fa -o output_dir/

Find more matches by harmonizing phased variants using refine:

   truvari refine output_dir/

Use Truvari’s comparison engine to consolidate redundant variants in a merged multi-sample VCF:

    bcftools merge -m none sampleA.vcf.gz sampleB.vcf.gz | bgzip > merge.vcf.gz
    tabix merge.vcf.gz
    truvari collapse -i merge.vcf.gz -o truvari_merge.vcf

🧬 Truvari Commands

• bench - Performance metrics from comparison of two VCFs
• collapse - Collapse possibly redundant VCF entries
• refine - Automated bench result refinement with phab
• anno - Add SV annotations to a VCF
• phab - Harmonize variant representations using MSA
• consistency - Consistency report between multiple VCFs
• vcf2df - Turn a VCF into a pandas DataFrame
• segment - Normalization of SVs into disjointed genomic regions
• stratify - Count variants per-region in vcf
• divide - Divide a VCF into independent shards
• ga4gh - Consolidate benchmarking result VCFs

🔎 More Information

All documentation about Truvari is on the WIKI. Additional information about using Truvari can be found in Discussions