@DIB-LAB/kProcessor

➤ Introduction

kProcessor is a C++ API with a handy Python interface that enables easy handling of sequence kmer content. kProcessor stores kmers with their associated metadata in a virtual data structure called kDataframe. By default, kDataFrame stores the kmers with their counts in the input dataset. However, kDataFrame supports adding multiple other columns to store more information about these kmers in different data types. Users can easily merge multiple kDataFrames or apply different set functions (e.g. union, intersect and difference) on a group of kDataFrames. Colored kDataFrame is another core virtual data structure in kProcessor that allows the indexing of the kmers in a multi-sequence reference input. It is composed of a kDataFrame that replaces the kmer count with a key (aka color). This key connects the kmer to all sequences associated with this kmer.

➤ Quick Installation (pip)

python -m pip install kProcessor

➤ Build from source

Clone

git clone https://github.com/dib-lab/kProcessor.git
cd kProcessor/
git submodule update --init --recursive

Install dependencies

sudo apt-get install g++ swig cmake python3-dev zlib1g-dev libghc-bzlib-dev python3-distutils libboost-all-dev

Build

CMake options

Build The kProcessor Library

# Run CMake configure
cmake -Bbuild

# Run make with parallel execution.
cmake --build build -j4 # -j4 = execute 4 recipes simultaneously.

Build Everything

cmake -Bbuild -DBUILD_ALL=1

cmake --build build -j4 # -j4 = execute 4 recipes simultaneously.

cmake -Bbuild -DBUILD_DOCS=1
cmake --build build --target GenerateDocs

cmake -Bbuild -DBUILD_TESTS=1
cmake --build build -j4

# Run tests
cd build/tests/kProcessorLibTests
./testKprocessorLib

➤ Manually build the Python bindings

Python bindings are generated using SWIG. It's recommended to install swig=4.0.2 using Conda.

You can build the python bindings by executing build_wrapper.sh, or you can follow the next steps.

Generate bindings

1. First, you need to follow the instructions in the Build from source.
2. While pwd=kProcessor run: python setup.py bdist_wheel.
3. Install the generated wheel package using: cd dist && python -m pip install kProcessor*.whl.

➤ Contributors

Mostafa Shokrof	Mohamed Abuelanin	Tamer Manosur

➤ License

Licensed under BSD-3-Clause.