SBPC - Simple Bayesian Peak Caller

SBPC is a peak caller for genomic data that implements a Bayesian statistical framework to evaluate the probabilities of regions being peaks. It is designed for CUT&TAG/CUT&RUN sequencing data and provides posterior probabilities for each detected peak region.

Usage

sbpc -b <sample>.bam -c <control>.bam > <sample>_peaks.bed

Installation

Pre-built Binaries

Linux

curl -L https://github.com/jakevc/sbpc/releases/download/v0.1.3/sbpc-v0.1.3-x86_64-unknown-linux-gnu.tar.gz | tar xz
sudo mv sbpc /usr/local/bin/

macOS (Intel)

curl -L https://github.com/jakevc/sbpc/releases/download/v0.1.3/sbpc-v0.1.3-x86_64-apple-darwin.tar.gz | tar xz
sudo mv sbpc /usr/local/bin/

macOS (Apple Silicon)

curl -L https://github.com/jakevc/sbpc/releases/download/v0.1.3/sbpc-v0.1.3-aarch64-apple-darwin.tar.gz | tar xz
sudo mv sbpc /usr/local/bin/

Using Cargo

cargo install sbpc

From Source

git clone https://github.com/jakevc/sbpc.git
cd sbpc
cargo build --release

The binary will be available at target/release/sbpc.

Using Docker

You can also run SBPC using Docker, which avoids the need to install system dependencies:

# Pull the latest image
docker pull ghcr.io/jakevc/sbpc:latest

# Run SBPC with Docker
docker run --rm -v /path/to/data:/data ghcr.io/jakevc/sbpc:latest -b /data/sample.bam -c /data/control.bam -o /data/output_prefix

This approach eliminates issues with glibc dependencies across different platforms, making it easier to use SBPC regardless of your operating system.

Requirements

SBPC requires input BAM files to be coordinate-sorted and indexed. You can prepare your BAM files using samtools:

samtools sort -o sorted.bam input.bam
samtools index sorted.bam

Both the sample and control BAM files must be sorted and have a corresponding .bai index file in the same directory.

Command Line Arguments

Usage: sbpc [OPTIONS] --bam <BAM>

Options:
  -b, --bam <BAM>              Input BAM file (must be paired-end reads)
  -c, --control <CONTROL>      Input BAM file with control signal to be normalized (e.g. IgG, Input)
  -s, --chromsize <CHROMSIZE>  Chromosome sizes for the genome if not found in the bam header
  -m, --mdist <MDIST>          Merge peaks within <mdist> base pairs [default: 1000]
  -r, --minreads <MINREADS>    Test genome bins with at least <minreads> read pairs [default: 15]
  -p, --posterior-threshold <THRESHOLD>  Define posterior probability threshold for peak significance [default: 0.95]
  -t, --step <STEP>            Bin size for coverage bins [default: 200]
  -w, --minwidth <MINWIDTH>    Minimum width (bp) of a peak [default: 150]
  -v, --version                Print the current SBPC version
      --broad                  Run SBPC on broad marks (--step 5000)
      --verbose                Run SBPC in verbose mode
  -h, --help                   Print help

Output

• Peaks are always printed in BED format to stdout.
• Metrics are printed as JSON to stdout by default. To also write metrics to a file (e.g., _sbpc.json), use the --metrics-file flag.

License

MIT