Conway-Bromage-Lyndon

A Rust library providing fully dynamic sets of k-mers with high locality.

The data structure is described in Conway-Bromage-Lyndon (CBL): an exact, dynamic representation of k-mer sets, please cite it if you use this library.

It supports the following operations:

• inserting a single k-mer (with insert), or every k-mer from a sequence (with insert_seq)
• deleting a single k-mer (with remove), or every k-mer from a sequence (with remove_seq)
• membership of a single k-mer (with contains), or every k-mer from a sequence (with contains_seq)
• iterating over the k-mers stored in the set (with iter)
• union / intersection / difference of two sets (with | / & / -)
• (de)serialization with serde

Requirements

Rust nightly 1.77+

If you have not installed Rust yet, please visit rustup.rs to install it. This library uses some nightly features of the Rust compiler (version 1.77+), you can install the latest nightly version with

rustup install nightly

If you don't want to use the +nightly flag every time you run cargo, you can set it as default with

rustup default nightly

Additional headers for Linux

This library uses C++ bindings for the sux library and tiered vectors. Depending on your configuration, some headers used for the bindings might be missing, in that case please install the following packages:

Ubuntu

sudo apt install -y libstdc++-12-dev libclang-dev

Fedora

sudo dnf install -y clang15-devel

Using the library

You can add CBL in an existing Rust project with

cargo +nightly add --git https://github.com/imartayan/CBL.git

or by adding the following dependency in your Cargo.toml

cbl = { git = "https://github.com/imartayan/CBL.git" }

If the build fails, try to install additional headers.

Choosing the right parameters

The CBL struct takes two main parameters as constants:

• an integer K specifying the size of the k-mers
• an integer type T (e.g. u32, u64, u128) that must be large enough to store both a k-mer and its number of bits together

Therefore T should be large enough to store $2k + \lg(2k)$ bits. In particular, since primitive integers cannot store more than 128 bits, this means that K must be ≤ 59.

Additionally, you can specify a third (optional) parameter PREFIX_BITS which determines the size of the underlying bitvector. Changing this parameter affects the space usage and the query time of the data structure, see the paper for more details.

Example usage

use cbl::CBL;
use needletail::parse_fastx_file;
use std::env::args;

// define the parameters K and T
const K: usize = 25;
type T = u64; // T must be large enough to store $2k + \lg(2k)$ bits

fn main() {
    let args: Vec<String> = args().collect();
    let input_filename = args.get(1).expect("No argument given");

    // create a CBL index with parameters K and T
    let mut cbl = CBL::<K, T>::new();

    let mut reader = parse_fastx_file(input_filename).unwrap();
    // for each sequence of the FASTA/Q file
    while let Some(record) = reader.next() {
        let seqrec = record.expect("Invalid record");

        // insert each k-mer of the sequence in the index
        cbl.insert_seq(&seqrec.seq());
    }
}

Building from source

You can clone the repository and its submodules with

git clone --recursive https://github.com/imartayan/CBL.git

If you did not use the --recursive flag, make sure to load the submodules with

git submodule update --init --recursive

Running the binaries

You can compile the binaries with

cargo +nightly build --release --examples

If the build fails, try to install additional headers.

By default, the binaries are compiled with a fixed K equal to 25, you can compile them with a different K as follows

K=59 cargo +nightly build --release --examples

Note that K values ≥ 60 are not supported by this library.

Similarly, PREFIX_BITS is equal to 24 by default and you can change it with

K=59 PREFIX_BITS=28 cargo +nightly build --release --examples

Note that PREFIX_BITS values ≥ 29 are not supported by this library.

Once compiled, the main binary will be located at target/release/examples/cbl. It supports the following commands:

Usage: cbl <COMMAND>

Commands:
  build        Build an index containing the k-mers of a FASTA/Q file
  count        Count the k-mers contained in an index
  list         List the k-mers contained in an index
  query        Query an index for every k-mer contained in a FASTA/Q file
  insert       Add the k-mers of a FASTA/Q file to an index
  remove       Remove the k-mers of a FASTA/Q file from an index
  merge        Compute the union of two indexes
  inter        Compute the intersection of two indexes
  diff         Compute the difference of two indexes
  sym-diff     Compute the symmetric difference of two indexes
  repartition  Show the repartition of the k-mers in the data structure
  help         Print this message or the help of the given subcommand(s)

Options:
  -h, --help     Print help
  -V, --version  Print version

Running the tests

You can run all the tests with

cargo +nightly test --lib

Building the documentation

You can build the documentation of the library and open it in your browser with

cargo +nightly doc --lib --no-deps --open

Citation

Conway-Bromage-Lyndon (CBL): an exact, dynamic representation of k-mer sets. Martayan, I., Cazaux, B., Limasset, A., and Marchet, C. https://doi.org/10.1093/bioinformatics/btae217

@article{cbl,
  title   = {{Conway–Bromage–Lyndon (CBL): an exact, dynamic representation of k-mer sets}},
  author  = {Martayan, Igor and Cazaux, Bastien and Limasset, Antoine and Marchet, Camille},
  journal = {Bioinformatics},
  volume  = {40},
  number  = {Supplement_1},
  pages   = {i48-i57},
  year    = {2024},
  month   = {06},
  issn    = {1367-4811},
  doi     = {10.1093/bioinformatics/btae217},
  url     = {https://doi.org/10.1093/bioinformatics/btae217},
  eprint  = {https://academic.oup.com/bioinformatics/article-pdf/40/Supplement\_1/i48/58354678/btae217.pdf}
}