bio-crispr-screens-base-editing-analysis

What it does

Quantifies per-sgRNA base edits, bystander rates, and indel byproducts from amplicon sequencing; maps sgRNAs to intended SNVs (target plus bystanders); computes per-variant fitness; annotates variants against ClinVar / COSMIC; provides decision rules for BE editor choice and screen design. It also includes procedures for filtering by editing efficiency, deconvoluting bystander contributions, and aggregating sgRNA scores to per-variant scores. Includes references to tools for BE analysis: CRISPResso --base_editor_output, Broad be-validation-pipeline, BE-Hive, and BE-Designer.

When to use it: Use during BE variant-screen design, choosing CBE vs ABE for a specific edit, interpreting bystander-confounded hits, distinguishing functional signal from indel artifact, integrating CRISPResso2 output with screen scoring, or deciding BE vs PE for SNV installation.

How it works

1. Base Editing Screen Analysis: For each sgRNA, quantify target-base conversion, bystander edits, and indel byproducts from amplicon sequencing; filter sgRNAs by editing efficiency; associate each sgRNA with its intended SNV (target plus bystander pattern).
2. Editor Chemistry Selection: Provides a table of editors with their windows and indel byproduct rates to guide BE3/BE4, ABE7.10/ABE8.x, eA3A-BE3, evoCDA-BE, CGBE1, GBE, and window extensions (e.g., ABE8e) usage.
3. Editing Window Math: Describes how editing windows (e.g., positions 4-8; 4-10 for ABE8e) affect which bases are edited; notes bystander confounding when multiple editable bases fall within the window.
4. sgRNA Library Design: Recommends tiling spacers so the SNV-of-interest base sits in the editing window with minimal bystander bases; annotates spacers with predicted amino acid changes (target + bystanders).
5. Editing Efficiency Filtering: Filters sgRNAs by target-editing percentage (e.g., >50%); cites thresholds from Hanna 2021 and notes context-specific adjustments.
6. Bystander Edit Attribution: Describes strategies to deconvolute target vs bystander contributions, including multiple sgRNAs with different bystander patterns, orthogonal chemistry, and bystander-stratified analysis.
7. Hit Calling: Filters by editing efficiency, then uses MAGeCK MLE or drugZ to score sgRNA-level results and aggregates to per-variant scores, distinguishing target-only from mixed-signal sgRNAs.
8. Case Studies: Summarizes Hanna 2021 BRCA1/2 variant-function screen methodology and Cuella-Martin 2021 DDR-gene screening as context for applying BE approaches.
9. Comparative Installation: Provides a comparison matrix Cas9/HDR, Cas9, CBE, ABE, CGBE/GBE, and Prime Editor for variant installation and bystander/indel considerations.

When to use it

Use when designing a BE variant screen, choosing CBE vs ABE for a specific edit, interpreting bystander-confounded hits, distinguishing functional signal from indel artifact, integrating CRISPResso2 output with screen scoring, or deciding BE vs PE for SNV installation.

What it can touch

It references and coordinates use of these tools and pipelines: CRISPResso (CLI: CRISPResso --base_editor_output), Broad be-validation-pipeline, BE-Hive (Python), BE-Designer (Python). It also implies integration with pandas, numpy, scipy, scikit-learn, and MAGeCK MLE / drugZ for downstream analysis.

Caveats

Versions and compatibility notes specify testing CRISPResso2 2.2.14+ and BE-Hive 1.0+, pandas 2.2+, biopython 1.83+, numpy 1.26+, scipy 1.12+, scikit-learn 1.4+, Broad be-validation-pipeline 1.0+. No explicit licensing beyond NOASSERTION; be-validation-pipeline and other tool usage should follow their respective licenses. The skill emphasizes that bystander edits can confound interpretation and that filtering thresholds are context-dependent, with cited benchmarks from Hanna 2021 and related studies.