Bigtools

Rust, CLI
Python

Bigtools is a library and associated tools for reading and writing bigwig and bigbed files.

The primary goals of the project are to be

• Performant
• Extensible
• Modern

Performant

Bigtools uses async/await internally to allow for efficient, multi-core computation when possible. In addition, tools are optimized for minimal memory usage. See Benchmarks for more details.

Extensible

Bigtools is designed to be as modular as possible. This, in addition to the safety and reliability of Rust, allows both flexibility and correctness as a library. In addition, its extremely easy to quickly create new tools or binaries. A number of binaries are available that parallel related existing binaries from UCSC, with drop-in compatibility for the most common flags.

Modern

Bigtools is written in Rust and published to crates.io, so binaries can be installed with cargo install bigtools or it can be used as a library by simply including it in your cargo.toml.

Library

To use bigtools in your Rust project, add bigtools to your Cargo.toml or run:

cargo add bigtools

See the bigtools 🦀 Documentation.

Example

use bigtools::bbiread::BigWigRead;

let mut reader = BigWigRead::open("test.bigWig").unwrap();
let chr1 = reader.get_interval("chr1", 0, 10000).unwrap();
for interval in chr1 {
    println!("{:?}", interval);
}

Binaries

The bigtools CLI binaries can be installed through crates.io or conda.

cargo install bigtools

conda install -c bioconda bigtools

Additionally, pre-built binaries can be downloaded through Github releases.

The following binaries are available:

Renaming the bigtools binary to any of the subcommands (case-insensitive) allows you to run that subcommand directly.

Python wrapper

The pybigtools package is a Python wrapper written using PyO3. It can be installed or used as a dependency either through PyPI or conda.

pip install pybigtools

conda install -c bioconda pybigtools

See the pybigtools 🐍 API Documentation.

How to build from source

In order to build the bigtools binaries, you can run

cargo build --release

and the binaries can be found in target/release/.

Otherwise, you can install the binaries from source by running

cargo install --path bigtools/

Building the python wheels for pybigtools requires maturin. To build the pybigtools wheel for installation (and install), you can run

maturin build --release -m pybigtools/Cargo.toml
pip install target/wheels/pybigtools*.whl

or

maturin develop --release -m pybigtools/Cargo.toml

Benchmarks

Benchmarks are included in the ./bench directory. They require python to run.

Multiple tools are compared against the comparable UCSC tools. For completeness, both single-threaded and multi-threaded (when available) benchmarks are included. Multiple different configuration options are benchmarked across multiple replicates, but a summar is available in the table below:

How to cite

This repository contains contains a CITATION.cff file with citation information. Github allows you to get a citation in either APA or BibTeX format; this is available in "Cite this repository" under About.