README.md

crisproff_jl

This is a Julia implementation for crisproff. For more details on the algorithm, please see their official repository.

Usage

Prerequisites

• Julia
• JLD2
• BioSequences
• TranscodingStreams
• CodecZlib
• Fire
• FASTX
• ViennaRNA_jll

You can download Julia from julialang and install these dependencies by simply typing commands like:

## type  ] to enter Pkg mode 
add JLD2

method

In order to process tons of gRNAs, the Julia version comes up here. So the software only supports FASTA sequences as input. In addition, I also think that you already use Risearch2 to search the genome to find all off-targets.

julia CRISPRspec_CRISPRoff_pipeline.jl --guides <guides.fa> --risearch-results-folder <folder> --CRISPRoff-scores-folder <folder> --specificity-report <file>

You can find the usage by:

julia CRISPRspec_CRISPRoff_pipeline.jl --help

Thanks for Julia language, the pipeline can be used at multi-theads mode, you can add -t <threads> after julia.

julia -t <Threads> CRISPRspec_CRISPRoff_pipeline.jl --guides <guides.fa> --risearch-results-folder <folder> --CRISPRoff-scores-folder <folder> --specificity-report <file>

Finally

It should have higher performance than python version! For 553 gRNAs, python version need run 2189s to get the results. For 553 gRNAs, julia version need run 1311s to get the results.but need 10 threads.

So, Julia version is 40% faster than python version when it use 10 threads! Julia version is not faster enough than python because of the most cost of time is from reading off-targets from Gzip file!

Citation

If you find this software useful for your research, please cite the following work:

CRISPR-Cas9 off-targeting assessment with nucleic acid duplex energy parameters Alkan F, Wenzel A, Anthon C, Havgaard JH, Gorodkin J Genome Biol. 2018 Oct 26;19(1):177

contact

xiaoguang.pan@bio.ku.dk