Predicts off-target propensity of CRISPR guide sequences

CRISPR Off-target Predictor (CROP) is a program which predicts the off-target propensity of a CRISPR guide sequence (guide) by creating combinations of substitutions, deletions and insertions for up to four positions along the guide 20 nucleotides. All of these combinations will then be mapped into the genome by using Bowtie (Langmead et al, 2009). The resulting mapped guide variants are counted to give the respective guide's off-target propensity. Two off-target scoring systems, namely Hsu (Hsu et al, 2013) and CFD (Doench et al, 2014; Doench et al, 2016) are also used to calculate the off-target score of each guide. This program takes an input of a folder containing gene sequences if (multi)-fasta formatted file and a reference genome for mapping purpose. It can take multiple genes at once but be aware that the number of input fasta files corresponds to the number of computer cores used.

Make sure that the reference genome your-genome.fasta has already been indexed using Bowtie (ref). To make an index:

1. place the genome file your-genome.fasta in genome/ directory
2. open bash and install bowtie: sudo apt install bowtie
3. navigate into genome/ directory and type: bowtie-build your-genome.fasta
4. the indexing will take several minutes depending on your computer. This process will gives you six files, namely: your-genome.1.ebwt, your-genome.2.ebwt, your-genome.3.ebwt, your-genome.4.ebwt, your-genome.rev.1.ebwt and your-genome.rev.1.ebwt.

1. Place all the fasta-formatted DNA sequences into one directory. Name this dirctory without using underscore character (e.g. input-folder/)
2. Open bash and navigate to crop/ directory. There are two .sh files, core.sh and crop.sh. Make both files into executables using chmod +x core.sh and chmod +x crop.sh, respectively
3. Run./crop.sh followed by entering input folder and your-genome
4. CROP is now running. Depending on your computer, it takes aroung 100 seconds to process one guide sequence. The total duration will follow the input sequence with most guides

CROP outputs the result named input-folder_out/. Each fasta file inside the input-folder/ were reported in input-folder_out/ as separate folders. Each folder contains:

1. _align/ containing align_map_s*.csv, a table containing all mapped variants aligned to the original guide sequence, comparing mismatch positions
2. _ot_maps/ containing map_s*.txt, a sam-formatted file listing mapped guide variants
3. _outfiles/ containing <seq>_out_doench.txt, a raw 30nt-guide sequence (N4-guide-NGG-N3) file as input for Azimuth score calculation at https://crispr.ml website; <seq>_out_guide_list.fasta, a fasta file containing all guide sequences extracted from <seq>; <seq>_score.csv, a table containing multiple informations for each guide, including GC content, Doench on-target score, map count, Hsu & CFD off-target score, and number of on target count
4. _single_guides/ containing <seq>.fasta, single-fasta files containing 20nt guide sequence concatenated with 78nt tracrRNA sequence. Both sequences are in RNA form (T > U)
5. <seq>_log.txt, a log file recording all the process taken to give the result and the time needed to finish the process

Included software:

• Bowtie: Artistic License 2.0

1. Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, (2009).
2. Hsu, P. D. et al. DNA targeting specificity of RNA-guided Cas9 nucleases. Nat. Biotechnol. 31, 827–832 (2013).
3. Doench, J. G. et al. Rational design of highly active sgRNAs for CRISPR-Cas9-mediated gene inactivation. Nat. Biotechnol. 32, 1262–1267 (2014).
4. Doench, J. G. et al. Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9. Nat. Biotechnol. 34, 184–191 (2016).