A Benchmark of Computational CRISPR-Cas9 Guide Design Methods Jacob Bradford, Dimitri Perrin. 2019.
This directory contains the following files (in alphabetical order):
(see each Python script for more information)
This script takes a .bed file and extracts desired portions as according to the configuration (adjusted within the script itself).
This scripts compares the guides generated by one tool, with the guides generated by every other tool.
A tab-separated values file containing annotation data for our 1m dataset (equivalent to: GRCm38/mm10 chromosome 19 10000000-11000000).
A tab-separated values file containing annotation data for our 500k dataset (equivalent to: GRCm38/mm10 chromosome 19 10000000-10500000).
A tab-separated values file containing annotation data for our 5m dataset (equivalent to: GRCm38/mm10 chromosome 19 10000000-15000000).
A tab-separated values file containing annotation data for our full dataset (equivalent to: GRCm38/mm10 chromosome 19).
The script used to generate the artificial genome, which contained 1169 experimentally validated guides.
The -TUSCAN edition of this script produces sequences which are 30 bp in length, rather than 23 bp.
The setup process for each tool is described in this RMarkdown file (see the HTML version for a rendered edition).
The targets which mm10db accepted when tested against the 500k dataset.
The targets which mm10db rejected when tested against the 500k dataset.
This scripts calculates the portion of guides reported by one tool, which would have been rejected by mm10db.
This script converts the raw data produced by the tested guide design tools into our standard format.
This script reads the normalised data (generated by normalise.py) and extracts guides that target exons.
The intentions of this script is to normalise the data produced in the tests using the experimentally validated dataset.
Generates a plot. The plot describes what percentage of exon-targeting guides reported by a particular tool would have been rejected by mm10db. The data for this is calculated by mm10db-rejects-accepted-by-other-tools.py.
This script generates the tool consensus heatmap plots in the paper.
This script takes the normalised (with scores, generated by normalise-Xu2015.py) data and plots it to visualise the recall and accuracy of each tool.
A modified file obtained from (see below), containing the experimentally validated guides. An additional column Xu-2015_Is-Efficient was appended to the data, which indicates whether the guide was determined to be efficient (1) or inefficient (0).
The data is formatted as: Gene Symbol,chrom,start of target,strand,sequence(target+3'+5'),"log2 fold change, HL60","log2 fold change, KBM7",Xu-2015_Is-Efficient
Xu, H., Xiao, T., Chen, C. H., Li, W., Meyer, C., Wu, Q., ... & Brown, M. (2015). Sequence determinants of improved CRISPR sgRNA design. Genome research, gr-191452.