crisprScore integration (optional, R)

Since v0.2 every on-target scorer is built into the Python package, so the Bioconductor crisprScore R bridge is optional – reach for it only for methods not ported to Python: RuleSet1, Azimuth, Lindel, CRISPRscan, CRISPRater, and the DeepHF T7 variants.

It is a packaged R wrapper, crisprscore_multi.R, run in a separate R/conda env that has the crisprScore package installed:

conda activate <crisprscore_r_env>

crisprscore_multi.R input.tsv <method_numbers> output_scored.tsv <enzyme> <5p_flank> <3p_flank>
# RuleSet1 + Azimuth + CRISPRater on Cas9 guides:
crisprscore_multi.R input.tsv 1,2,11 scored.tsv Cas9 13 29

Input is a TSV with a context column (everything else is preserved); the script trims context to each method’s required length. Methods:

#	Method	Enzyme
1	RuleSet1	Cas9
2	Azimuth	Cas9
3-8	DeepHF variants (U6 / T7)	Cas9
9	Lindel	Cas9
10	CRISPRscan	Cas9
11	CRISPRater	Cas9
12	DeepSpCas9	Cas9
13-14	RuleSet3 (Hsu / Chen)	Cas9
15-16	DeepCpf1	Cas12a
17	EnPAMGB	Cas12a

Note

Many of these (12, 13-17, DeepHF U6) are already native in score_guides and don’t need R. Use crisprScore mainly for 1, 2, 9, 10, 11.