Published January 17, 2025 | Version v3

PhEval: A Framework for Evaluating Phenotype-Driven Variant and Gene Prioritisation Algorithms

  • 1. ROR icon Queen Mary University of London
  • 2. ROR icon European Bioinformatics Institute
  • 3. ROR icon University of Colorado Anschutz Medical Campus
  • 4. EDMO icon University of North Carolina Chapel Hill
  • 5. ROR icon Lawrence Berkeley National Laboratory
  • 6. ROR icon National and Kapodistrian University of Athens
  • 7. Semanticly
  • 8. Wellcome Trust Sanger Institute
  • 9. ROR icon Charité - Universitätsmedizin Berlin

Description

This repository contains the complete corpora and PhEval processed TSV results referenced in the study "PhEval: A Framework for Evaluating Phenotype-Driven Variant and Gene Prioritisation Algorithms". The provided materials are integral to the benchmarking process and include:

  1. Phenopacket Corpora:

    • Phenotype-Only Dataset: Phenopacket store containing phenotypic data without genomic or structural variant information.
    • Phenotype and Genomic Variant Dataset: Phenopacket store with phenotypic data and associated genomic variants.
    • Phenotype and Structural Variant Dataset: Structural variant corpus combining phenotypic data with structural variant information.
  2. PhEval TSV Results:

    • TSV files summarizing the outcomes of nine different benchmarking configurations tested in the paper, including:
      • Phenotype-Only Tools: Exomiser (phenotype-only mode), Phen2Gene, PhenoGenius, and GADO.
      • Phenotype and Genomic Variant Tools: Exomiser (with VCF), LIRICAL, and AI-MARRVEL.
      • Structural Variant Tools: Exomiser and SvAnna.

Files

Files (5.6 GB)

Name Size
md5:0bebaaa74fa92d02110d1a71f53a0699
5.7 MB Download
md5:ace8c0fc7b886f3f31bac6df462e89b7
5.6 GB Download