whatshap

TAFFISH wrapper for WhatsHap 2.8. WhatsHap phases genomic variants from sequencing reads and provides related haplotype-aware utilities such as polyploid phasing, haplotagging, phasing statistics, phasing comparison, and read splitting by haplotype.

This app packages the upstream whatshap command from Bioconda and keeps TAFFISH command mode enabled. It also includes samtools 1.23.1 and bcftools 1.23.1 as practical BAM/VCF helpers for indexing, compression, and small format checks.

Usage

Show the TAFFISH wrapper help:

taf-whatshap --help

Show upstream WhatsHap help:

taf-whatshap -- --help
taf-whatshap -- --version

Run diploid read-based phasing:

taf-whatshap whatshap phase \
  --reference reference.fa \
  -o phased.vcf \
  variants.vcf \
  alignments.bam

Run haplotagging on an indexed phased VCF/BCF and an indexed alignment:

taf-whatshap whatshap haplotag \
  --reference reference.fa \
  -o haplotagged.bam \
  phased.vcf.gz \
  alignments.bam

Generate phasing statistics:

taf-whatshap whatshap stats --tsv stats.tsv phased.vcf.gz

Because TAFFISH command mode is enabled, helper executables from the same container can also be called directly:

taf-whatshap samtools --version
taf-whatshap bcftools --version
taf-whatshap bgzip --help
taf-whatshap tabix --help

For WhatsHap subcommands, the most portable form is taf-whatshap whatshap SUBCOMMAND ..., which explicitly runs the upstream whatshap executable inside the container. The direct forms taf-whatshap phase ... and taf-whatshap -- phase ... are interpreted by TAFFISH command mode as attempts to run an executable named phase inside the container. Use -- mainly for option-leading arguments to the default command, such as taf-whatshap -- --help.

Packaged Commands

• whatshap
• samtools
• bcftools
• bgzip
• tabix
• python

The main WhatsHap subcommands in 2.8 are compare, find_snv_candidates, genotype, hapcut2vcf, haplotag, haplotagphase, learn, phase, polyphase, polyphasegenetic, split, stats, and unphase.

Inputs and Outputs

Typical inputs are:

• VCF/BCF variant files with genotypes.
• BAM/CRAM alignment files, usually coordinate-sorted and indexed.
• Reference FASTA for variant types beyond the simplest SNV/indel cases, CRAM input, haplotagging, and robust allele detection.
• Optional pedigree or technology-specific inputs for the corresponding upstream subcommands.

Typical outputs are:

• Phased VCF/BCF with genotype phasing and phase-set annotations.
• Haplotagged BAM with haplotype tags from whatshap haplotag.
• Statistics tables and reports from whatshap stats or related commands.
• Split read files from whatshap split.

Exact options and output formats are controlled by upstream WhatsHap. Use taf-whatshap whatshap SUBCOMMAND --help for subcommand-specific details.

Boundaries

This app does not perform read alignment, variant calling, BAM sorting, automatic reference download, or full workflow orchestration. It exposes the upstream WhatsHap command line in a pinned, containerized TAFFISH environment. Input quality, variant representation, phasing informativeness, and read length/coverage strongly affect scientific results.

The image is intended to run offline after installation. It does not download reference genomes or external databases at runtime.

Platform

The image is built for native linux/amd64 and linux/arm64. WhatsHap 2.8, samtools 1.23.1, and bcftools 1.23.1 are installed from Bioconda/conda-forge with architecture-specific compiled dependencies.

Metadata

• Upstream: https://github.com/whatshap/whatshap
• Documentation: https://whatshap.readthedocs.io/
• Upstream version: 2.8
• TAFFISH release: 2.8-r1
• TAFFISH app license: Apache-2.0
• Upstream license: MIT
• Citation: Martin et al. 2016. WhatsHap: fast and accurate read-based phasing. DOI: 10.1101/085050