pafchainer

A tool for processing PAF (Pairwise Alignment Format) files and merging alignment chains using the Wavefront Alignment Algorithm (WFA).

Overview

pafchainer merges consecutive alignments in a chain by connecting them with wavefront alignments. It erodes the boundaries of adjacent alignments and fills in the gaps with new alignments, creating a single, contiguous alignment path. This is particularly useful for genomic alignment workflows that require complete end-to-end alignments.

Features

Processes PAF files with chain information
Uses the Wavefront Alignment Algorithm for fast and accurate gap alignments
Indexes chains for efficient processing
Handles both compressed (BGZF) and uncompressed PAF files
Configurable boundary erosion size
Outputs in PAF or SAM format
Multi-threaded processing

Installation

Clone the repository and build using Cargo:

git clone https://github.com/waveygang/pafchainer
cd pafchainer
cargo build --release

Usage

pafchainer --paf input.paf --query query.fa --target target.fa [OPTIONS]

Required Arguments:

--paf, -p: Input PAF file with chain information
--query, -q: Query FASTA file
--target, -t: Target FASTA file

Optional Arguments:

--erosion-size, -e: Size of boundary erosion in base pairs (default: 100)
--output, -o: Output file (default: stdout)
--sam: Output in SAM format instead of PAF
--threads, -t: Number of threads to use (default: 4)
--verbose, -v: Verbosity level (0=error, 1=info, 2=debug)

Examples

Basic usage:

pafchainer -p alignments.paf -q query.fa -t reference.fa -o connected.paf

Output in SAM format:

pafchainer -p alignments.paf -q query.fa -t reference.fa --sam -o aligned.sam

How It Works

Builds or loads a chain index for efficient access to PAF entries
For each chain, erodes boundaries between adjacent entries
Performs wavefront alignment to connect the eroded regions
Merges CIGAR strings to create a single continuous alignment
Outputs the merged alignment in PAF or SAM format

License

MIT License

Pafchainer

Install / Use

README