Public CRISPR guide counting: Yusa/MAGeCK lane

• Category: Public data

• Assay type: crispr

• Evidence level: Documented example

• Dataset: MAGeCK/Yusa public CRISPR guide-counting FASTQs

• Condition shown: Known-good guide-counting lane with count agreement and Edlib validation.

• Primary report: docs/benchmarks/public_crispr/README.md

What This Shows

• DotMatch can run the documented fixed-window CRISPR guide-counting workflow on a public guide-counting lane.

• The report shows exact, Hamming, and Levenshtein lanes separately instead of mixing their semantics.

• Count-agreement and Edlib-validation artifacts are available for inspection.

What It Does Not Show

• This is not a universal CRISPR-screen analysis claim.

• Downstream essentiality, differential abundance, and screen interpretation remain outside DotMatch.

Comparator And Validation

Comparator semantics: DotMatch guide counts are checked against public CRISPR workflow rows, count-agreement artifacts, and native Edlib assignment validation as documented in the benchmark report.

Validation: The public CRISPR gate requires repeated rows, positive throughput records, count agreement artifacts, and zero Edlib assignment mismatches for checked reads.

Report Examples

• Public CRISPR benchmark report (markdown)

• Count agreement summary (csv)

Raw Artifacts

• benchmarks/raw/public_crispr_repeated.csv

• benchmarks/raw/public_crispr_edlib_validation.csv

• benchmarks/raw/count_agreement_summary.csv

Reproduce Or Check

• make bench-public-crispr-repeated

• make public-crispr-evidence-gate