Hasan: Haplotype Analysis from BAM files

Hasan (Haplotype Algorithm for SNP Amplicon Networks) is a Python package for analyzing haplotypes from BAM files using SNP information. It constructs directed acyclic graphs (DAGs) to identify and visualize potential haplotypes based on sequencing data.

Features

• Read and process SNP information from TSV files
• Convert VCF files to compatible TSV format
• Build phasing tables from BAM files
• Create directed acyclic graphs (DAGs) for haplotype visualization
• Find and analyze potential haplotypes
• Interactive graph visualization with draggable nodes
• Command-line interface with rich output formatting

Installation

pip install hasan

Requirements

• Python ≥ 3.6
• pysam
• pandas
• networkx
• matplotlib
• click
• rich

Usage

Command Line Interface

The package provides two main commands:

1. Analyze haplotypes:

hasan analyze <bam_file> <snps_file> [options]

Options:

• --method/-m: Method for calculating path weights ('min' or 'multiply')
• --plot/--no-plot: Enable/disable interactive plot visualization
• --output/-o: Specify output TSV file for haplotype results
• --verbose/-v: Print detailed progress information

Example:

hasan analyze sample.bam variants.tsv --method min --plot --output results.tsv --verbose

2. Convert VCF to TSV:

hasan convert <input_vcf> <output_tsv> [options]

Options:

• --verbose/-v: Print detailed progress information

Example:

hasan convert variants.vcf variants.tsv --verbose

Input File Formats

SNPs File (TSV format)

CHROM   POS     REF     ALT     QUAL    DP
chr1    1000    A       G       40      20
chr1    1500    C       T       35      15

Note: When converting from VCF, variants are filtered to:

• Exclude indels (only SNPs are kept)
• Require minimum quality score (QUAL ≥ 30)
• Require minimum depth (DP ≥ 10)

Python API

from hasan import read_snps, build_phasing_table, create_dag, find_haplotypes

# Read SNP information
snps_df = read_snps("variants.tsv")

# Build phasing table
phasing_data = build_phasing_table("sample.bam", snps_df)

# Create graph
G = create_dag(phasing_data, snps_df)

# Find haplotypes
haplotypes = find_haplotypes(G, method='min')

Output

The package provides multiple output formats:

1. Interactive visualization (when using --plot)
2. Static graph image (haplotype_graph.png)
3. TSV file with haplotype frequencies (when using --output)
4. Rich console output showing:
   • Haplotype sequences
   • Proportions for each haplotype
   • Progress information (in verbose mode)

How It Works

1. SNP Reading: Loads SNP positions and variants from a TSV file.
2. Phasing Table: Processes BAM file to count base occurrences at SNP positions.
3. Graph Construction: Creates a DAG where:
   • Nodes represent bases at each position
   • Edges represent connections between consecutive positions
   • Edge weights represent proportion of reads supporting the connection
4. Haplotype Finding: Identifies possible haplotypes by finding paths through the graph.

Visualization

The interactive visualization allows you to:

• Drag nodes to rearrange the graph
• View edge weights representing read proportions
• Distinguish between reference (green) and alternate (blue) bases