Last Updated: April 21, 2026 | Focus: pre-order CRISPR library sizing, guide QC, and synthesis-ready submission checks
Design and QC a CRISPR sgRNA Library Before Ordering
Use this workflow when the guide set needs to become an orderable CRISPR library instead of staying a loose design spreadsheet. The core job is to size the library, clean the guide set, confirm coverage and representation assumptions, and prepare a synthesis-ready submission without carrying obvious dropout or formatting risk into the vendor handoff.
Need the adjacent vendor-choice page?
Stay on this page if the library still needs QC, coverage planning, or synthesis-ready cleanup. Switch to the CRISPR synthesis-platform page if the guide set is already stable and the next decision is vendor fit.
Quick Takeaways
- •Treat coverage planning, guide QC, and synthesis formatting as one submission workflow rather than isolated checks.
- •The fastest way to avoid a bad CRISPR pool order is to clean dropout and skew risk before the vendor file exists.
- •Genome-wide, pathway, and tiling libraries fail for different reasons, so the QC emphasis has to match the library type.
- •Vendor submission is the last step, not the place to discover guide-set instability or missing controls.
Prerequisites
Required Knowledge:
- Basic CRISPR/Cas9 principles and experimental design
- sgRNA design software (e.g., Benchling, GPP Web Portal, CRISPRscan)
- Target gene list or genomic regions for screening
Required Files:
- sgRNA sequences (20-23 bp, typically 20 bp for SpCas9)
- Target gene IDs and genomic coordinates
- Scaffold sequence (constant region, e.g., tracrRNA)
Tools Needed:
- sgRNA design tool (not provided by OligoPool.com - use Benchling/GPP/etc.)
- OligoPool.com tools: Coverage Calculator, GC Analyzer, Batch QC
- Spreadsheet software (Excel/Google Sheets) for library compilation
CRISPR Library Types & Cas9 Systems
Cas9 System Specifications
| System | PAM | Guide Length | Application |
|---|---|---|---|
| SpCas9 | 5'-NGG-3' | 20 bp (17-24 bp) | Standard knockout, most common |
| SaCas9 | 5'-NNGRRT-3' | 21 bp | AAV delivery (smaller size) |
| Cas12a (Cpf1) | 5'-TTTV-3' | 23-25 bp | AT-rich targets, multiplexing |
| dCas9-VP64 | 5'-NGG-3' | 20 bp | CRISPRa (activation) libraries |
* Use GC Content Analyzer to verify guide compatibility with chosen Cas9 system
Genome-Wide Knockout
Target all protein-coding genes (18,000-20,000 genes). Phenotype discovery.
- • Guides/gene: 4-6 (knockout), 10 (CRISPRa/i)
- • Total sgRNAs: 72,000-200,000
- • Controls: 1,000 non-targeting + essential genes
- • Example: Human GeCKO v2 (123K guides)
Targeted/Pathway Libraries
Gene families, pathways, or drug targets (100-1,000 genes). Focused validation.
- • Guides/gene: 5-10 (higher redundancy)
- • Total sgRNAs: 500-10,000
- • Use case: Kinome, GPCRs, epigenetics
- • Advantage: Deep coverage, affordable
Tiling/Regulatory
Dense sgRNA coverage of non-coding regions. Enhancer/promoter mapping.
- • Density: 1 guide every 5-10 bp
- • Size: Depends on locus (1-100 kb typical)
- • Use case: CRISPRi of enhancers/promoters
- • Tool: Use Coverage Calculator for tiling design
Design Workflow
Calculate Library Coverage
Determine how many sgRNAs you need using the Coverage Calculator.
Instructions:
- Go to Coverage Calculator
- Select "CRISPR Library" mode
- Enter number of target genes (e.g., 18,000 for genome-wide human)
- Set guides per gene (typically 4-6 for knockout, 3-10 for activation)
- Include non-targeting controls (recommend 500-1,000 guides)
- Review total oligo count and estimated cost
Typical Library Sizes:
| Library Type | Genes | sgRNA/Gene | Total sgRNAs |
|---|---|---|---|
| Human Genome-Wide | ~19,000 | 4-6 | 76,000-114,000 |
| Mouse Genome-Wide | ~22,000 | 4-6 | 88,000-132,000 |
| Targeted (Kinases) | ~500 | 10 | 5,000 |
| CRISPRa/i | ~18,000 | 10 | 180,000 |
* Add 1,000 non-targeting controls to any library
Design Guidelines:
- Redundancy: 4-6 sgRNAs/gene for robust phenotypes
- Controls: Include non-targeting and essential gene controls
- Power: Ensure sufficient coverage for statistical analysis (1000x representation)
Design sgRNA Sequences
Use validated design algorithms to generate guides with high predicted activity and minimal off-targets.
Design Software (External):
OligoPool.com provides QC tools only. Use these for sgRNA design:
- Algorithm: Doench 2016 (Rule Set 2)
- Score range: 0-100
- Free, web-based
- Broad Institute, validated libraries
- Human/mouse genome-wide
- Includes Azimuth scores
- ML-based, latest algorithm
- Multi-species support
- Off-target scoring
- Moreno-Mateos 2015 algorithm
- Command-line, batch processing
- Good for zebrafish/mouse
sgRNA Design Criteria (Quantitative):
Activity Score Interpretation:
Include metadata in FASTA headers for traceability. Filter by Doench >0.5 and MIT >50 before QC.
QC sgRNA Sequences
Screen all sgRNA sequences for quality issues before adding scaffolds.
A. GC Content Analysis
- Navigate to GC Content Analyzer (batch mode)
- Upload FASTA file with all sgRNA spacer sequences (20 bp only)
- Flag sequences: <25% GC (poor activity) or >75% GC (synthesis issues)
- Export flagged list, design replacements for critical targets
Impact on Efficiency:
GC 40-60%: optimal activity. GC <25%: significantly reduced cutting efficiency (30-50% in published studies). GC >75%: increased synthesis failure risk (15-25% reported by array vendors). See How to Analyze GC Content Across Oligo Pools.
B. Batch Sequence QC
- Go to Batch Sequence QC
- Upload FASTA with all sgRNA spacers
- Flag homopolymers: GGGG (high dropout risk), TTTT (U6 termination), AAAA (synthesis errors)
- Check for palindromes and tandem repeats (synthesis artifacts)
Rejection Criteria:
- TTTT or longer: Mandatory removal (pol-III termination)
- GGGG runs: High library dropout risk (≥40-50% observed in NGS QC)
- Low complexity (<1.5 Shannon entropy): Replace if critical gene
C. Secondary Structure Check
- Use Secondary Structure Predictor for flagged sequences
- Upload sgRNA spacer + scaffold (full 96 bp construct)
- Check ΔG < -3 kcal/mol for stable hairpins in spacer region
- Prioritize replacement if hairpin blocks PAM-proximal region (positions 1-12)
Hairpin Impact:
Stable hairpins (ΔG < -3 kcal/mol) in seed region (bp 13-20): significant activity reduction (40-60% reported). See Secondary Structure Tutorial.
Important: Sequences to Remove:
- 1. Poly-T (TTTT or longer) → causes transcription termination
- 2. Extreme GC (<20% or >80%)
- 3. Perfect palindromes → may form hairpins
- 4. Sequences matching your cloning vector
Replacement Strategy:
For each removed guide, design an alternative targeting nearby region (±50 bp). Maintain the same number of guides per gene for balanced representation.
Add Scaffold & Cloning Adapters
Assemble full-length oligos with system-specific scaffolds and vector-compatible adapters.
Oligo Structure by Cas9 System:
SpCas9 (lentiCRISPR v2 / pLKO):
SaCas9 (pX601):
CRISPRa (SAM MS2 system):
Oligo Assembly in Excel/Python:
- Create spreadsheet with columns: Gene, sgRNA_ID, spacer_sequence
- Add columns for: 5'-adapter, scaffold, 3'-adapter (constant for all)
- Concatenate: =CONCATENATE(adapter_5, spacer, scaffold, adapter_3)
- Add barcode column (optional, for deconvolution)
- Export as CSV or FASTA for synthesis order
Important Considerations:
- Strand: Order sense strand (same as U6 transcription)
- Format: Check vendor requirements (single vs. double-stranded)
- Length limit: Most arrays support up to 230 bp (you'll be ~100-110 bp)
- Barcoding: Add unique barcode if multiplexing libraries
Final QC & Synthesis Order
Perform final quality checks on assembled oligos before placing synthesis order.
Final Checklist:
Synthesis Vendor Options:
- Scale: 1K-300K oligos
- Length: 60-300 bp
- Cost: $0.04-0.08/oligo (scale-dependent)
- NGS QC: Included (500×)
- Lead time: 2-3 weeks
- Scale: Up to 1M features
- Length: 60-200 bp
- Cost: $0.03-0.06/oligo (100K+ pools)
- QC: Optional NGS add-on
- Lead time: 3-4 weeks
- Scale: 12K-2M oligos
- Length: 40-350 bp
- Cost: $0.08-0.12/oligo (mid-scale)
- High uniformity (<3-fold CV)
- Lead time: 2-3 weeks
* Pricing as of Q4 2026, varies by library size and specifications. Compare vendors: Oligo Pool Synthesis Vendors
NGS QC Metrics (Post-Synthesis):
File Format for Synthesis Order:
Option 1: FASTA Format (Recommended)
Header format: >unique_ID|gene_name|guide_number|metadata (optional)
Option 2: CSV/TSV Format
Required columns: oligo_id (unique), sequence (full-length oligo). Optional: gene, metadata.
- Twist: FASTA or CSV, max 300 bp, no special characters in IDs
- Agilent: Tab-delimited text file, oligo name + sequence columns
- CustomArray: Excel template provided, FASTA also accepted
Post-Synthesis Steps (Brief):
- Resuspend oligo pool in TE buffer
- PCR amplify with minimal cycles (6-8 cycles)
- Clone into lentiviral vector (Gibson assembly or restriction cloning)
- Transform, grow overnight, maxi-prep
- NGS verification of library representation
- Package lentivirus and transduce target cells
Detailed cloning protocols are beyond the scope of this calculator guide.
Post-Screen Data Analysis
After completing your CRISPR screen, use statistical algorithms to identify significant hits from NGS count data.
Analysis Tools
- MAGeCK: Most popular, RRA algorithm, gene-level FDR
- BAGEL: Bayesian approach, best for essentiality screens
- drugZ: Optimized for drug resistance/synthetic lethality
- RIGER: Early method, rank-based, still widely used
Hit Calling Thresholds
- FDR: <0.05 (stringent), <0.25 (permissive for validation)
- Log2 Fold-Change: |Δ| >1.5 (depletion/enrichment screens)
- Guide concordance: ≥3/6 guides significant per gene
- Effect size: Consider biological relevance, not just p-value
Statistical Considerations:
Best Practices
Design Redundancy
Always include 4-6 guides per gene. Statistical power for hit calling requires multiple independent guides showing consistent phenotypes.
Include Controls
1,000 non-targeting guides (negative control) + essential gene guides (positive control, e.g., POLR2A, RPA3) for quality assessment.
Plan Sequencing Depth
Aim for 500-1000x coverage per sgRNA in NGS QC. For 100K library, that's 50-100 million reads. Budget for deep sequencing.
Validate Library Quality
ALWAYS perform NGS on final plasmid library before virus production. Check for representation, dropouts, and skew. Aim for <10% dropout.
Troubleshooting Library Issues
Problem: High dropout rate (>20% of designed guides)
Diagnose by dropout pattern:
- 20-40% dropout, random: Synthesis failure. Check for poly-T (TTTT+), extreme GC (<25%, >75%), or secondary structures. Redesign flagged sequences.
- 20-30% dropout, sequence-specific: PCR amplification bias. Reduce cycles from 12→6-8, use KAPA HiFi or Q5 polymerase.
- >40% dropout: Transformation bottleneck. Need ≥10× library complexity as CFUs (e.g., 100K library = 1M colonies minimum). Pool 5-10 transformations.
- Dropout enriched in high-GC guides: PCR bias during amplification. Switch to long-extension polymerase (1 min/kb).
Problem: Skewed representation (CV >5-fold)
Optimize PCR amplification:
- Reduce cycles: Start with 6 cycles, increase only if yield <1 µg. Each extra cycle progressively increases skew (typically ~10-15% additional bias observed).
- Template amount: Use 10-100 ng input DNA. Too low (<1 ng) = stochastic sampling. Too high (>500 ng) = incomplete denaturation.
- Polymerase choice: KAPA HiFi (lowest bias), Q5 (good), Phusion (moderate bias). Avoid Taq (extreme GC bias).
- Alternative: Emulsion PCR or linear amplification (IVT) for highly uniform libraries.
Problem: Low cloning efficiency (<105 CFU/µg)
Systematic optimization:
- Vector prep: High-quality maxi-prep, A260/280 = 1.8-1.9. Digest overnight, gel-purify, dephosphorylate (CIP or rSAP).
- Insert:vector ratio: Start 1:5 molar ratio, optimize to 1:10 or 1:20 if low efficiency.
- Competent cells: Use ≥109 CFU/µg cells (Endura, Stbl3, or ElectroMAX). Test efficiency with control plasmid before library.
- Transformation scale: For 100K library, perform 10-20 transformations of 10 µL cells each. Pool after recovery.
- Recovery: 1 hour at 37°C in SOC medium before plating. Do NOT exceed 1.5 hours (some guides may replicate faster).
Problem: Sequences present in input pool but lost after cloning
Likely toxic or unstable in E. coli:
- Cryptic promoters: Some sgRNA sequences may form bacterial promoters. Use recombination-deficient strains (Stbl3).
- Palindromes: Inverted repeats can trigger recombination. These should have been flagged in Step 3 QC.
- Solution: Clone in low-copy plasmid (pSC101 origin) or switch to yeast-based library construction (more stable but complex).
Workflow Summary
Coverage Calculation
sgRNA Design
Sequence QC
Add Scaffolds
Order & Validate
Quick Reference: CRISPR Library Design Parameters
For experienced users — critical thresholds at a glance.
| Parameter | Optimal Range | Reject If | Tool |
|---|---|---|---|
| sgRNA Length | 20 bp (SpCas9) | <17 bp or >24 bp | Design software |
| GC Content | 40-60% | <25% or >75% | GC Analyzer |
| Activity Score | Doench >0.5, Azimuth >50 | Doench <0.3 | Benchling, GPP |
| Off-Target CFD | MIT score >50 | CFD aggregate >0.20 | CRISPick |
| Poly-T Runs | 0 (absent) | ≥TTTT (U6 termination) | Batch QC |
| Hairpin ΔG | >-2 kcal/mol | <-3 kcal/mol in seed | Structure Predictor |
| Guides per Gene | 4-6 (knockout), 10 (CRISPRa) | <3 (low power) | Coverage Calculator |
| Non-Targeting Controls | 500-1,000 guides | <100 | Manual design |
| Oligo Length | 100-110 bp (total) | >230 bp (array limit) | Spreadsheet |
| NGS Coverage (QC) | 500-1000× per guide | <100× (insufficient) | Vendor QC report |
| Library Dropout | <10% | >20% (redesign needed) | NGS analysis |
| Representation Skew | <3-fold CV | >5-fold (PCR bias) | NGS analysis |