Phasing Pipeline

A Nextflow pipeline for phasing unphased genotype data using Beagle with reference panels from the 1000 Genomes Project.

Overview

This pipeline performs haplotype phasing of unphased VCF files using Beagle, a powerful tool for phasing and imputation. The pipeline includes preprocessing steps to prepare both the unphased data and reference panels, performs chromosome-by-chromosome phasing, and postprocesses the results.

Features

• Preprocessing: Quality control and preparation of unphased VCF files
• Reference Panel Processing: Indexing and preparation of 1000 Genomes reference panels
• Phasing: Chromosome-by-chromosome phasing using Beagle
• Postprocessing: Concatenation and final processing of phased results
• Containerized: Uses Singularity containers for reproducibility
• Scalable: Supports SLURM cluster execution

Requirements

• Nextflow (>= 22.04.0)
• Singularity (for containerized execution)
• SLURM (for cluster execution, optional)

Quick Start

1. Clone the repository:

   git clone <repository-url>
   cd Phasing

2. Configure parameters: Edit params/params_beagle.yml with your input files:

   vcf_unphased: ./data/your_unphased.vcf.gz
   refcsv: ./params/files_beagle.csv
   outdir: ./results/
   output_prefix: phased_output

3. Run the pipeline:

   # Local execution
   nextflow run main.nf -params-file params/params_beagle.yml -profile local
   
   # SLURM cluster execution
   nextflow run main.nf -params-file params/params_beagle.yml -profile kutral

Input Files

Required

• Unphased VCF: A VCF/BCF file containing unphased genotypes (must be indexed)
• Reference CSV: A CSV file with columns:
  • chr: Chromosome identifier
  • ref_vcf: Path to reference VCF file for that chromosome
  • ref_vcf_index: Path to reference VCF index file
  • gmap: Path to genetic map file

Example Reference CSV (files_beagle.csv):

chr,ref_vcf,ref_vcf_index,gmap
1,/path/to/chr1_ref.vcf.gz,/path/to/chr1_ref.vcf.gz.csi,/path/to/chr1.gmap
2,/path/to/chr2_ref.vcf.gz,/path/to/chr2_ref.vcf.gz.csi,/path/to/chr2.gmap
...

Pipeline Workflow

1. Preprocessing:

   • Index VCF files
   • Fill AC (allele count) annotations
   • Remove duplicate variants
   • Remove missing genotypes
   • Prepare reference panels

2. Phasing:

   • Extract chromosome-specific regions
   • Run Beagle phasing with reference panels
   • Index phased output

3. Postprocessing:

   • Concatenate chromosome-specific results
   • Generate final phased VCF

Output

The pipeline generates:

• Phased VCF files per chromosome: phased_<chr>.vcf.gz
• Concatenated phased VCF: <output_prefix>.vcf.gz
• Pipeline execution reports in pipeline_info/

Configuration

Profiles

• local: For local execution with Singularity
• kutral: For SLURM cluster execution on the ngen-ko queue

Resource Requirements

Default resource allocation:

• Memory: 120GB per process
• CPUs: 16 per process

Adjust in nextflow.config if needed.

Parameters

Key parameters (set in params/params_beagle.yml):

Parameter	Description	Default
vcf_unphased	Path to unphased VCF file	-
refcsv	Path to reference CSV file	-
outdir	Output directory	./results/
output_prefix	Prefix for output files	-

Tools Used

• Beagle: Haplotype phasing and imputation
• bcftools: VCF/BCF manipulation and indexing
• Nextflow: Workflow orchestration

Citation

If you use this pipeline, please cite:

• Beagle: Browning, B. L., & Browning, S. R. (2016). Genotype imputation with millions of reference samples. The American Journal of Human Genetics, 98(1), 116-126.

License

See LICENSE file for details.

Author

Gabriel Cabas

Support

For issues and questions, please open an issue on the repository.