Overview

kb-hashing is a locality sensitive hashing tool for genetic sequence comparison in bioinformatics.

Compile kb-hashing

Use the following to compile kb-hashing:

cd src
g++ -o kbhashing kbhashing.cpp -std=c++11

Simulation data generation

We provide a data simulation program and corresponding script to generate simulated data. Use the following to compile it and run the script:

cd src
g++ -o seqsim seqsim.cpp -std=c++11
cd ../src/script
./simulation.sh n m

Here n means the length of generated sequences and m represents the number of pairs of sequence in each data set. The generated data will be under the folder data. There will be six datasets in that folder with different error rate from 0.05 to 0.3.

Usage

For kbhashing, you use the following commands to run the program:

./src/kbhashing k b m l t < inputfile

Here all the parameters are introduced in our paper. The inputfile must be some pairs of sequences only consist of ACGT. And each line should only has one sequence.

We also provide a script of pipeline to get the results in our paper. You need to generate the simulation data before running the pipeline.

cd ./script
./simulation.sh n m
./pipeline.sh k b m l t n m