slivar

slivar filters variant call format (VCF) files using JavaScript expressions to prioritize single-nucleotide (SNP) and insertion/deletion (INDEL) candidate variants for rare-disease genetic analyses.


Key Features:

  • Flexible filtering: Filter variants on genotype quality, sequencing depth, allele balance, and population allele frequency.
  • JavaScript expression engine: Apply complex, user-defined filters through simple JavaScript expressions evaluated against VCF records.
  • Inheritance-model support: Support filtering strategies for de novo, recessive, dominant, compound-heterozygous, autosomal recessive, and X-linked variants.
  • Trio and group analysis: Operate on trios and groups to identify inheritance patterns and candidate variants within pedigrees.
  • Large-scale dataset performance and expected yields: Optimized for large genomic datasets with reported yields of ~10 candidate SNP/INDELs per exome and ~19 per genome, including averages of ~3 de novo, ~10 compound-heterozygotes, ~1 autosomal recessive, ~4 X-linked, and ~100 autosomal dominant candidates per whole genome.
  • VCF integration: Integrate filters and resulting annotations into VCF files to produce filtered candidate lists.

Scientific Applications:

  • Familial rare-disease studies: Prioritize candidate causal variants in trios and pedigrees to support genotype–phenotype investigations.
  • Inheritance-model analyses: Detect and prioritize de novo, compound-heterozygous, autosomal recessive, autosomal dominant, and X-linked candidate variants.
  • Variant-filtering guideline development: Derive and apply filtering thresholds based on genotype quality, depth, allele balance, and population allele frequency to produce reproducible candidate lists.

Methodology:

Apply JavaScript expressions to VCF genotype and annotation fields (e.g., genotype quality, sequencing depth, allele balance, population allele frequency), derive filtering guidelines from these attributes, and integrate the filters into VCF files to produce candidate variant lists with reported expected yields.

Topics

Collections

Details

License:
MIT
Tool Type:
command-line tool
Programming Languages:
JavaScript, Python, Shell
Added:
1/20/2021
Last Updated:
5/20/2021

Operations

Publications

Pedersen BS, Brown JM, Dashnow H, Wallace AD, Velinder M, Tvrdik T, Mao R, Best DH, Bayrak-Toydemir P, Quinlan AR. Effective variant filtering and expected candidate variant yield in studies of rare human disease. Unknown Journal. 2020. doi:10.1101/2020.08.13.249532.