CRISPR Screen Guider Tutorial

Start with the study layer

Important: a variant project is automatically locked to Prime Editing MAVE. It never falls back to sgRNA/gene-level library presets, spacer-only QC, or gene-level analysis.

Run sample data and inspect the result

On Library QC, choose a modality and click Load example beside the FASTQ input. Run QC, then compare assigned reads, missing reagents, and Gini or count CV. These metrics describe plasmid-library representation, not screen hits.

On Analyze, load the matching T0, endpoint, and manifest examples. Run the comparison and inspect replicate agreement before the ranked gene or variant effects. A coherent direction across multiple reagents supports interpretation; the sample data are a software demonstration, not independent experimental validation.

The same input fields accept your own files. Sample FASTQ and manifest files can be downloaded next to each Load example control to check the required format.

View variant-level sample output View gene-level sample output

Variant-level Prime Editing MAVE: practical path

1. PLAN: choose variant-level functional screen. Enter expected variant count and pegRNAs per variant only for scale calculations; do not upload variants here.
2. BUILD: upload a CSV, TSV, VCF, or pasted table with chrom,pos,ref,alt. Positions must be 1-based GRCh38 coordinates. The site verifies the reference allele before design.
3. DESIGN: run the local PRIDICT2 worker. CRISPR Screen Guider filters and validates candidates, selects Top 3 by default, uses the fixed 86-nt PE scaffold, and runs fixed-seed pegLIT linker engineering.
4. LIBRARY: build the vector-aware variable insert for sgOpti-tevoPQ1 (Addgene #255532): spacer → scaffold → RTT → PBS → linker. The vector contributes tevopreQ1 and 7T, so those sequences are not duplicated in the insert.
5. QC: upload plasmid FASTQ plus the Build-generated pegRNA manifest. Assignment requires the full identity signature, not a 20-nt spacer alone.
6. ANALYZE: upload T0 and endpoint FASTQ files plus the manifest. The site calculates pegRNA abundance effects and aggregates multiple pegRNAs to variants.
7. REVIEW: compare planned variants/pegRNAs with observed pegRNA/variant representation. V1 reports abundance-based effects only; it does not infer editing efficiency from PRIDICT2 or count abundance.

Open Prime Editing Build Read Prime Editing references

Gene-level CRISPR screen: practical path

1. PLAN: choose gene-level discovery, then select CRISPR-KO, CRISPRi, or CRISPRa. Set library size, MOI, representation, replicates, selection, gDNA, PCR, and sequencing assumptions.
2. BUILD: enter gene symbols and choose published guide sources. The output is an sgRNA library manifest and cloning package.
3. QC: upload plasmid FASTQ plus the sgRNA manifest. Review guide detection, missing guides, Gini coefficient, count CV, and gene completeness.
4. SCREEN: record T0, control, treatment, endpoint, and replicate structure.
5. ANALYZE: upload T0 and endpoint FASTQ files plus the shared guide manifest, then run MAGeCK-based sgRNA-to-gene analysis.
6. REVIEW: compare planned guides and targets with observed sgRNA and gene representation before interpreting hit genes.

Open gene-level Build Open gene-level Analyze

Inputs and terminology

References and scope

PRIDICT2 ranks pegRNA candidates, pegLIT designs linkers, and MAGeCK analyzes guide counts. See References. Prime Editing MAVE currently supports human GRCh38 SNVs with PE2-style NGG design. It does not support complex rearrangements, large insertions or deletions, PE3/PE5, twinPE, or PASSIGE.