🧠 CRISPR-Cas9 Fundamentals

Q: What is a CRISPR-Cas9 guide RNA?

A CRISPR-Cas9 guide RNA (gRNA) is a short synthetic RNA (~20 nucleotides) that directs the Cas9 nuclease to a specific DNA sequence complementary to the guide. The gRNA forms a complex with Cas9, which then binds and cleaves the target DNA adjacent to a PAM sequence.

Q: What is a PAM sequence?

A PAM (Protospacer Adjacent Motif) is a short DNA sequence (typically 5'-NGG-3' for SpCas9) that is required for Cas9 binding and cleavage. The PAM is located immediately 3' of the target sequence. Without a PAM, Cas9 cannot cleave the DNA.

Q: What is off-target effect in CRISPR?

Off-target effects occur when the guide RNA directs Cas9 to cleave unintended genomic sites that are similar (but not identical) to the target sequence. These can cause mutations elsewhere in the genome. Minimize off-targets by choosing guides with high specificity scores and verifying with tools like Cas-OFFinder.

🎯 What is CRISPR-Cas9?

CRISPR-Cas9 is a revolutionary gene-editing technology that allows precise modification of DNA sequences. Originally discovered as a bacterial immune system, it has been repurposed as a programmable nuclease. The system consists of two components: the Cas9 enzyme (which cuts DNA) and a guide RNA (gRNA) (which directs Cas9 to the target site). Guide RNA design is the critical first step — a good guide ensures high on-target efficiency and minimal off-target effects.

Why Guide Design Matters

Precision: Targets exact location. Efficiency: High cutting rate. Safety: Low off-targets. Specificity: Unique genome match. Function: Knockout, knock-in, editing. Reproducibility: Consistent results.

📌 gRNA📌 Cas9📌 PAM📌 Off-target

# CRISPR-Cas9 Guide Design — the big picture
# Designing guide RNAs for precise genome editing.

print("""
=== CRISPR-Cas9 Guide Design ===

What it is:
  - gRNA directs Cas9 to target
  - Cas9 cuts DNA
  - Programmable nuclease
  - Gene editing tool

Components:
  - Cas9 enzyme
  - Guide RNA (gRNA)
  - PAM sequence
  - Target DNA

Why guide design matters:
  - On-target efficiency
  - Off-target effects
  - Specificity
  - Reproducibility

Applications:
  - Gene knockout
  - Gene knock-in
  - Base editing
  - Prime editing
  - Transcriptional regulation
  - Epigenome editing
  - Disease modeling
  - Crop improvement
  - Gene therapy
""")

print("=== Historical Context ===")
print("""
1987: CRISPR discovered in E. coli
2005: CRISPR associated with immunity
2012: CRISPR-Cas9 repurposed for editing (Doudna & Charpentier)
2013: First mammalian cell editing
2015: CRISPR used in crops
2020: Nobel Prize in Chemistry
2023: First CRISPR therapy approved (Casgevy)
2025: Base editing, prime editing, AI-designed guides
""")

print("=== Why It's Hard ===")
print("""
1. OFF-TARGET EFFECTS
   - Similar sequences elsewhere
   - Unintended cuts
   - Safety concerns

2. ON-TARGET EFFICIENCY
   - Not all guides work
   - Variable efficiency
   - Context-dependent

3. PAM CONSTRAINT
   - NGG for SpCas9
   - Limits targeting
   - Alternative Cas9 variants

4. SECONDARY STRUCTURE
   - gRNA folding
   - Reduced efficiency
   - Hard to predict

5. CHROMATIN
   - Accessibility
   - Closed regions
   - Variable efficiency
""")

🎯 Why Guide Design Matters

🎯 The Value of Good Guides

A well-designed guide RNA is the difference between a successful experiment and a failed one. Guide design determines efficiency, specificity, and safety.

Key Factors

Efficiency: Cutting rate. Specificity: Unique target. Safety: Low off-target. Position: Functional domain. GC: 40-60%. Structure: Accessible.

🎯 Efficiency🎯 Specificity🎯 Safety🎯 Position

# Why Guide Design Matters
print("=== 1. On-Target Efficiency ===")
print("""
Efficiency:
  - Percentage of cells with edit
  - Varies from 0% to 100%
  - Guide-dependent

Factors:
  - GC content (40-60%)
  - No homopolymers (TTTT)
  - Position in gene
  - Chromatin accessibility
  - Secondary structure

Example:
  Guide A: 80% efficiency
  Guide B: 10% efficiency
  → Choose Guide A
""")

print("\n=== 2. Specificity ===")
print("""
Specificity:
  - Unique target in genome
  - No similar sequences
  - Low off-target

Factors:
  - Seed sequence (3' end)
  - Mismatch tolerance
  - PAM proximity
  - Genome complexity

Example:
  Guide A: 0 off-targets
  Guide B: 50 off-targets
  → Choose Guide A
""")

print("\n=== 3. Safety ===")
print("""
Safety:
  - Minimize off-target cuts
  - Avoid oncogenes
  - Avoid tumor suppressors
  - Clinical applications

Concerns:
  - Off-target mutations
  - Chromosomal rearrangements
  - Gene activation
  - Gene silencing

Example:
  Clinical trial
  → Only high-specificity guides
""")

print("\n=== 4. Position in Gene ===")
print("""
Knockout:
  - Early exon
  - Functional domain
  - Conserved region

Knock-in:
  - Near insertion site
  - HDR efficiency
  - Distance from cut

Example:
  Exon 1 (5' end)
  → Frameshift
  → Knockout
""")

print("\n=== 5. GC Content ===")
print("""
GC content:
  - 40-60% optimal
  - Too low: weak binding
  - Too high: secondary structure

Example:
  Guide A: 50% GC → good
  Guide B: 30% GC → poor
  Guide C: 80% GC → poor
""")

print("\n=== 6. Secondary Structure ===")
print("""
Secondary structure:
  - gRNA folding
  - Hairpins
  - Reduced efficiency

Avoid:
  - Strong hairpins
  - Self-complementarity
  - Stable structures

Example:
  Guide A: no structure → good
  Guide B: hairpin → poor
""")

print("\n=== 7. Chromatin Accessibility ===")
print("""
Chromatin:
  - Open regions → accessible
  - Closed regions → inaccessible
  - Cell-type specific

Tools:
  - ATAC-seq
  - DNase-seq
  - ChIP-seq

Example:
  Open chromatin → high efficiency
  Closed chromatin → low efficiency
""")

print("\n=== 8. Reproducibility ===")
print("""
Reproducibility:
  - Same guide → same result
  - Batch effects
  - Protocol

Factors:
  - Delivery method
  - Cell type
  - Time point

Example:
  Guide A in lab 1: 80%
  Guide A in lab 2: 75%
  → Reproducible
""")

📍 PAM Sequences

🎯 Protospacer Adjacent Motif

The PAM (Protospacer Adjacent Motif) is a short DNA sequence required for Cas9 binding and cleavage. The PAM is located immediately 3' of the target sequence. Without a PAM, Cas9 cannot cleave DNA.

PAM Essentials

SpCas9: NGG. SaCas9: NNGRRT. Cas12a: TTTV. Position: 3' of target. Required: For cleavage. Variant: Engineered PAMs.

📍 NGG📍 NNGRRT📍 TTTV📍 Variants

# PAM Sequences
print("=== PAM for Different Cas9 ===")
print("""
SpCas9 (Streptococcus pyogenes):
  - PAM: 5'-NGG-3'
  - Most common
  - 3' of target
  - N = any base

SaCas9 (Staphylococcus aureus):
  - PAM: 5'-NNGRRT-3'
  - Smaller Cas9
  - AAV delivery

Cas12a (Cpf1):
  - PAM: 5'-TTTV-3'
  - 5' of target
  - Staggered cut

Cas13:
  - RNA targeting
  - No PAM
  - Different mechanism

Engineered:
  - SpCas9-NG: NG PAM
  - SpRY: NRN/NYN
  - Near-PAMless
""")

print("\n=== PAM Position ===")
print("""
SpCas9:
  5'-[target 20nt]-NGG-3'
                  ^
                  PAM

Example:
  Target: GACGTACGTACGTACGTAC
  PAM: TGG
  Full: GACGTACGTACGTACGTACTGG

Cas12a:
  5'-TTTV-[target 20nt]-3'
    ^
    PAM
""")

print("\n=== PAM Finding ===")
print("""
Algorithm:
  1. Scan sequence
  2. Find NGG (or variant)
  3. Extract 20nt upstream
  4. Candidate guide

Example:
  Sequence: ...ACGTACGTACGTACGTACGTGG...
  Find: TGG at position 21
  Guide: ACGTACGTACGTACGTAC (20nt)
""")

print("\n=== PAM Variants ===")
print("""
SpCas9 variants:
  - Wild-type: NGG
  - SpCas9-VQR: NGA
  - SpCas9-EQR: NGAG
  - SpCas9-NG: NG
  - SpRY: NRN/NYN
  - Near-PAMless

Use cases:
  - Wild-type: standard
  - NG: targeting AT-rich
  - SpRY: flexible
""")

print("\n=== PAM Direction ===")
print("""
SpCas9:
  - Target: 5' to 3'
  - PAM: 3' of target
  - Guide: complementary to target

Example:
  Target: 5'-GACGTACGTACGTACGTAC-3'
  PAM: 3'-TGG-5' (on opposite strand)
  Guide: 5'-GACGUACGUACGUACGUAC-3' (RNA)
""")

print("\n=== Alternative PAMs ===")
print("""
1. SpCas9-NG
   - PAM: NG
   - Targets more sites
   - Reduced specificity

2. SpRY
   - PAM: NRN/NYN
   - Near-PAMless
   - Flexible targeting

3. SaCas9
   - PAM: NNGRRT
   - Smaller
   - AAV delivery

4. Cas12a
   - PAM: TTTV
   - 5' of target
   - Staggered cut

5. Cas12b
   - PAM: TTN
   - Alternative

6. CasX
   - PAM: TTC
   - Small
""")

print("\n=== PAM Design Rules ===")
print("""
1. Find PAM (NGG for SpCas9)
2. Extract 20nt upstream
3. Check GC content (40-60%)
4. Avoid homopolymers (TTTT)
5. Check off-targets
6. Score on-target
7. Check secondary structure
8. Order oligos
""")

⚙️ Guide Design Principles

🎯 Designing a Guide RNA

Good guide design follows a set of principles: appropriate GC content, no homopolymers, correct position, and minimal off-targets. Follow these rules to maximize success.

Design Principles

GC: 40-60%. Length: 20nt. Homopolymer: Avoid TTTT. Position: Early exon. Strand: Either. Off-target: Minimize.

⚙️ GC 40-60%⚙️ 20nt⚙️ No TTTT⚙️ Early exon

# Guide Design Principles
print("=== 1. GC Content ===")
print("""
Optimal: 40-60%
  - < 40%: weak binding
  - > 60%: secondary structure

Example:
  Guide A: GC = 50% → good
  Guide B: GC = 30% → poor
  Guide C: GC = 80% → poor
""")

print("\n=== 2. Length ===")
print("""
Standard: 20 nucleotides
  - SpCas9: 20nt
  - SaCas9: 21-23nt
  - Cas12a: 23nt

Truncated guides:
  - 17-18nt
  - Reduced off-target
  - Lower efficiency

Example:
  Full: 20nt
  Truncated: 18nt
""")

print("\n=== 3. Homopolymers ===")
print("""
Avoid:
  - TTTT (Pol III terminator)
  - AAAA
  - GGGG (G-quadruplex)
  - CCCC

Example:
  Guide A: no TTTT → good
  Guide B: TTTT → poor
""")

print("\n=== 4. Position ===")
print("""
Knockout:
  - Early exon (5' end)
  - Functional domain
  - Conserved region

Knock-in:
  - Near insertion site
  - HDR efficiency
  - Distance from cut

Example:
  Exon 1 (5' end)
  → Frameshift
  → Knockout
""")

print("\n=== 5. Strand ===")
print("""
Either strand:
  - Both work
  - No preference

Example:
  Forward strand: guide A
  Reverse strand: guide B
  → Both valid
""")

print("\n=== 6. Off-Target ===")
print("""
Minimize:
  - Unique seed sequence
  - Mismatch in seed
  - PAM-proximal mismatches

Seed region:
  - 3' end of guide
  - 8-12nt
  - Critical for specificity

Example:
  Guide A: unique seed → good
  Guide B: seed matches elsewhere → poor
""")

print("\n=== 7. Secondary Structure ===")
print("""
Avoid:
  - Hairpins
  - Self-complementarity
  - Stable structures

Check:
  - Mfold
  - RNAfold
  - CRISPRscan

Example:
  Guide A: no structure → good
  Guide B: hairpin → poor
""")

print("\n=== 8. Chromatin ===")
print("""
Accessibility:
  - Open regions → accessible
  - Closed regions → inaccessible
  - Cell-type specific

Tools:
  - ATAC-seq
  - DNase-seq
  - ChIP-seq

Example:
  Open chromatin → high efficiency
  Closed chromatin → low efficiency
""")

print("\n=== Design Workflow ===")
print("""
1. Define target region
2. Find PAMs (NGG)
3. Extract 20nt guides
4. Filter GC content (40-60%)
5. Remove homopolymers (TTTT)
6. Check off-targets
7. Score on-target
8. Check secondary structure
9. Select top 3-5 guides
10. Order oligos
""")

print("\n=== Example Design ===")
print("""
Target gene: BRCA1
Region: Exon 1
Sequence: ATGGATTTATCTGCTCTTCGCGTTGAAGAAGTACAAAATGTCATTAATGCTATGCAGAAAATCTTAGAGTGTCCCATCTGTCTGGAGTTGATCAAGGAACCTGTCTCCACAAAGTGTGACCACATATTTTGCAAATTTTGCATGCTGAAACTTCTCAACCAGAAGAAAGGGCCTTCACAGTGTCCTTTATGTAAGAATGATATAACCAAAAGGAGCCTACAAGAAAGTACGAGATTTAGTCAACTTGTTGAAGAGCTATTGAAAATCATTTGTGCTTTTCAGCTTGACACAGGTTTGGAGTATGCAAACAGCTATAATTTTGCAAAAAAGGAAAATAACTCTCCTGAACATCTAAAAGATGAAGTTTCTATCATCCAAAGTATGGGCTACAGAAACCGTGCCAAAAGACTTCTACAGAGTGAACCCGAAAATCCTTCCTTGCAGGAAACCAGTCTCAGTGTCCAACTCTCTAACCTTGGAACTGTGAGAACTCTGAGGACAAAGCAGCGGATACAACCTCAAAAGACGTCTGTCTACATTGAATTGGGATCTGATTCTTCTGAAGATACCGTTAATAAGGCAACTTATTGCAGTGTGGGAGATCAAGAATTGTTACAAATCA""")

print("=== Example Guides ===")
print("""
Guide 1: ATGGATTTATCTGCTCTTCG (GC=45%)
Guide 2: TGGATTTATCTGCTCTTCGC (GC=50%)
Guide 3: GGATTTATCTGCTCTTCGCG (GC=55%)
Guide 4: GATTTATCTGCTCTTCGCGT (GC=50%)
Guide 5: ATTTATCTGCTCTTCGCGTT (GC=45%)

Select top 3-5 based on:
  - GC content
  - Off-target score
  - On-target score
""")

🎯 On-Target Scoring

🎯 Predicting Efficiency

On-target scoring predicts how efficiently a guide will cut its intended target. Multiple algorithms exist, each trained on different datasets.

On-Target Scoring

Algorithms: CRISPRscan, Doench, Wang. Features: GC, position, structure. Score: 0-1. Threshold: >0.5 good. Context: Cell-type. Validation: Experimental.

🎯 CRISPRscan🎯 Doench🎯 Wang🎯 Score

# On-Target Scoring
print("=== CRISPRscan ===")
print("""
CRISPRscan:
  - Developed by Moreno-Mateos lab
  - For zebrafish, human, mouse
  - Based on in vivo data
  - Score 0-1

Features:
  - GC content
  - Position-specific
  - Secondary structure
  - Seed sequence

Use:
  - Score guides
  - Select top guides
  - Web: crisprscan.org
""")

print("\n=== Doench (Rule Set 2) ===")
print("""
Doench et al. (2016):
  - Broad Institute
  - Machine learning
  - Human/mouse cells
  - Score 0-1

Features:
  - GC content
  - Position-specific
  - Nucleotide composition
  - Secondary structure

Use:
  - Benchling
  - Broad GPP
  - Web tool
""")

print("\n=== Wang (SSC) ===")
print("""
Wang et al. (2014):
  - Sequence features
  - SVM
  - Human cells
  - Score 0-1

Features:
  - GC content
  - Position-specific
  - Dinucleotide composition

Use:
  - CRISPRscan
  - Web tool
""")

print("\n=== Scoring Features ===")
print("""
1. GC content
   - Optimal: 40-60%

2. Position-specific
   - Nucleotides at each position
   - Learned weights

3. Secondary structure
   - Free energy
   - Hairpins
   - Accessibility

4. Seed sequence
   - 3' end
   - Critical for specificity

5. Chromatin
   - Accessibility
   - Open/closed

6. Distance from PAM
   - Proximal vs distal
""")

print("\n=== Score Interpretation ===")
print("""
Score range: 0-1

Interpretation:
  > 0.7: Excellent
  0.5-0.7: Good
  0.3-0.5: Moderate
  < 0.3: Poor

Example:
  Guide A: 0.85 → excellent
  Guide B: 0.60 → good
  Guide C: 0.25 → poor
""")

print("\n=== Tools ===")
print("""
1. CRISPRscan
   - crisprscan.org
   - Zebrafish, human, mouse

2. Benchling
   - benchling.com
   - Doench score
   - Off-target

3. CHOPCHOP
   - chopchop.cbu.uib.no
   - Multiple scores

4. Broad GPP
   - portals.broadinstitute.org/gpp
   - Doench score

5. CRISPOR
   - crispor.tefor.net
   - Multiple scores
   - Off-target
""")

print("\n=== Python Scoring ===")
print("""
# Simple GC content score
def gc_score(guide):
    gc = (guide.count('G') + guide.count('C')) / len(guide)
    return 1 - abs(gc - 0.5) * 2  # 1 at 50% GC

# Example
guides = [
    "ATGGATTTATCTGCTCTTCG",
    "TGGATTTATCTGCTCTTCGC",
    "GGATTTATCTGCTCTTCGCG",
]

for g in guides:
    print(f"{g}: GC={gc_score(g):.2f}")
""")

print("\n=== Experimental Validation ===")
print("""
Validate:
  - T7 endonuclease assay
  - Surveyor assay
  - Deep sequencing
  - Amplicon sequencing

Efficiency:
  - Percentage of edited cells
  - Indel frequency
  - Knockout rate

Example:
  Guide A: 80% indels
  Guide B: 20% indels
  → Choose Guide A
""")

🚫 Off-Target Prediction

🎯 Finding Unintended Cuts

Off-target prediction identifies potential unintended cleavage sites in the genome. Minimizing off-targets is critical for safety and reproducibility.

Off-Target Prediction

Mismatches: Up to 3-5. Seed: 3' end critical. Bulges: DNA/RNA. Tools: Cas-OFFinder. Score: CFD, MIT. Validate: GUIDE-seq.

🚫 Mismatches🚫 Seed🚫 CFD🚫 Cas-OFFinder

# Off-Target Prediction
print("=== Off-Target Basics ===")
print("""
Off-target:
  - Unintended cleavage
  - Similar sequences
  - Mismatches allowed

Mismatch tolerance:
  - Up to 3-5 mismatches
  - Seed region (3' end) critical
  - PAM-proximal mismatches

Example:
  Target: GACGTACGTACGTACGTAC
  Off-target: GACGTACGTACGTACGTAA (1 mismatch)
  → Potential off-target
""")

print("\n=== Seed Sequence ===")
print("""
Seed:
  - 3' end of guide
  - 8-12nt
  - Critical for specificity

Mismatches:
  - Seed mismatch → no cleavage
  - Non-seed mismatch → may cleave

Example:
  Guide: GACGTACGTACGTACGTAC
  Seed: TACGTAC (last 7nt)
  → Critical
""")

print("\n=== CFD Score ===")
print("""
CFD (Cutting Frequency Determination):
  - Doench et al.
  - Score 0-1
  - Higher = more off-target

Features:
  - Mismatch position
  - Mismatch type
  - Seed region

Use:
  - Benchling
  - CRISPOR
  - Broad GPP
""")

print("\n=== MIT Score ===")
print("""
MIT Score:
  - Zhang lab
  - Score 0-100
  - Higher = more off-target

Features:
  - Mismatch position
  - Mismatch count
  - Seed region

Use:
  - CRISPRscan
  - Web tool
""")

print("\n=== Cas-OFFinder ===")
print("""
Cas-OFFinder:
  - Fast
  - Allows mismatches
  - Allows bulges
  - Genome-wide

Input:
  - Guide sequence
  - PAM
  - Genome
  - Mismatch threshold

Output:
  - Off-target sites
  - Mismatch count
  - Position

Command:
  cas-offinder guide.txt genome.fa out.txt
""")

print("\n=== Off-Target Tools ===")
print("""
1. Cas-OFFinder
   - Fast
   - Genome-wide
   - Mismatches/bulges

2. CRISPOR
   - Web
   - Multiple scores
   - Off-target

3. Benchling
   - Web
   - CFD score
   - Off-target

4. CHOPCHOP
   - Web
   - Off-target
   - Multiple genomes

5. FlashFry
   - Fast
   - Genome-wide
   - Command line

6. GUIDE-seq
   - Experimental
   - Unbiased
   - Genome-wide
""")

print("\n=== Minimizing Off-Targets ===")
print("""
1. Unique seed sequence
2. Mismatch in seed
3. PAM-proximal mismatches
4. Truncated guides (17-18nt)
5. High-fidelity Cas9
6. Nickase (paired)
7. Ribonucleoprotein delivery

Example:
  Guide A: 0 off-targets → good
  Guide B: 50 off-targets → poor
""")

print("\n=== High-Fidelity Cas9 ===")
print("""
Variants:
  - SpCas9-HF1
  - eSpCas9
  - HypaCas9
  - HiFi Cas9

Features:
  - Reduced off-target
  - Maintained on-target
  - Better specificity

Use:
  - Clinical
  - Sensitive applications
""")

print("\n=== Validation ===")
print("""
Experimental:
  - GUIDE-seq
  - DISCOVER-seq
  - CIRCLE-seq
  - SITE-seq

Computational:
  - Cas-OFFinder
  - CRISPOR
  - Benchling
""")

🧬 Secondary Structure

🎯 gRNA Folding

The secondary structure of the guide RNA can affect its ability to bind Cas9 and target DNA. Strong secondary structures reduce efficiency.

Secondary Structure

Hairpins: Avoid. Self-complementarity: Avoid. Free energy: Less negative. Accessibility: Seed region. Tools: RNAfold, Mfold. CRISPRscan: Score.

🧬 Hairpins🧬 Self-complement🧬 Free energy🧬 RNAfold

# Secondary Structure
print("=== Why Structure Matters ===")
print("""
Secondary structure:
  - gRNA folding
  - Hairpins
  - Self-complementarity
  - Reduced efficiency

Impact:
  - Cas9 binding
  - Target binding
  - Cleavage

Example:
  Guide A: no structure → good
  Guide B: hairpin → poor
""")

print("\n=== Hairpins ===")
print("""
Hairpin:
  - Stem-loop
  - Self-complementary
  - Stable
  - Blocks binding

Example:
  5'-GCGCGC...GCGCGC-3'
      |||||     |||||
      stem      loop

Avoid:
  - Strong stems
  - Long stems
  - GC-rich stems
""")

print("\n=== Self-Complementarity ===")
print("""
Self-complementarity:
  - Guide binds itself
  - Dimers
  - Reduced efficiency

Example:
  Guide: GACGTACGTACGTACGTAC
  Complement: CTGCATGCATGCATGCATG
  → Potential dimer

Avoid:
  - Long complementary regions
  - 3' complementarity
""")

print("\n=== Free Energy ===")
print("""
Free energy (ΔG):
  - More negative = more stable
  - Less negative = less structure

Optimal:
  - ΔG > -10 kcal/mol
  - Less stable

Example:
  Guide A: ΔG = -5 → good
  Guide B: ΔG = -20 → poor
""")

print("\n=== Accessibility ===")
print("""
Accessibility:
  - Seed region (3' end)
  - Must be accessible
  - Not base-paired

Check:
  - RNAfold
  - Mfold
  - CRISPRscan

Example:
  Seed accessible → good
  Seed paired → poor
""")

print("\n=== Tools ===")
print("""
1. RNAfold
   - ViennaRNA
   - Free energy
   - Structure

2. Mfold
   - Web
   - Structure
   - Free energy

3. CRISPRscan
   - Score
   - Structure
   - Accessibility

4. RNAstructure
   - Free energy
   - Structure
""")

print("\n=== Python Structure ===")
print("""
# Simple hairpin check
def has_hairpin(guide):
    # Check for self-complementarity
    complement = {'A':'T', 'T':'A', 'G':'C', 'C':'G'}
    rev = guide[::-1]
    comp = ''.join(complement.get(b, 'N') for b in rev)
    
    # Check for matches
    matches = sum(1 for a, b in zip(guide, comp) if a == b)
    return matches > 5  # threshold

# Example
guides = [
    "ATGGATTTATCTGCTCTTCG",
    "GCGCGCGCGCGCGCGCGCGC",
]

for g in guides:
    print(f"{g}: hairpin={has_hairpin(g)}")
""")

print("\n=== Design Rules ===")
print("""
1. Avoid hairpins
2. Avoid self-complementarity
3. Check free energy
4. Check seed accessibility
5. Use CRISPRscan score
6. Validate experimentally
""")

🛠️ Tools (CRISPRscan, Benchling, CHOPCHOP)

🎯 Guide Design Tools

Many tools exist for CRISPR guide design. Each has strengths and is suited to different use cases.

Tools

CRISPRscan: In vivo. Benchling: All-in-one. CHOPCHOP: Free. CRISPOR: Comprehensive. Broad GPP: Doench. Cas-OFFinder: Off-target.

🛠️ CRISPRscan🛠️ Benchling🛠️ CHOPCHOP🛠️ CRISPOR

# Tools
print("=== CRISPRscan ===")
print("""
CRISPRscan:
  - Web: crisprscan.org
  - In vivo data
  - Zebrafish, human, mouse
  - Score 0-1
  - Free

Features:
  - On-target score
  - Off-target
  - Secondary structure
  - Visualization

Input:
  - Target sequence
  - PAM
  - Genome

Output:
  - Guides
  - Scores
  - Off-targets
""")

print("\n=== Benchling ===")
print("""
Benchling:
  - Web: benchling.com
  - All-in-one
  - Free for academics
  - Doench score
  - Off-target

Features:
  - Guide design
  - On-target score
  - Off-target
  - Plasmid design
  - Notebook

Input:
  - Target gene
  - Genome

Output:
  - Guides
  - Scores
  - Off-targets
""")

print("\n=== CHOPCHOP ===")
print("""
CHOPCHOP:
  - Web: chopchop.cbu.uib.no
  - Free
  - Multiple genomes
  - Multiple scores

Features:
  - On-target score
  - Off-target
  - Multiple PAMs
  - Visualization

Input:
  - Target gene
  - Genome
  - PAM

Output:
  - Guides
  - Scores
  - Off-targets
""")

print("\n=== CRISPOR ===")
print("""
CRISPOR:
  - Web: crispor.tefor.net
  - Comprehensive
  - Multiple scores
  - Off-target

Features:
  - On-target score
  - Off-target
  - Multiple tools
  - Visualization

Input:
  - Target sequence
  - Genome

Output:
  - Guides
  - Scores
  - Off-targets
""")

print("\n=== Broad GPP ===")
print("""
Broad GPP:
  - Web: portals.broadinstitute.org/gpp
  - Doench score
  - Off-target
  - CRISPRko, CRISPRi, CRISPRa

Features:
  - On-target score
  - Off-target
  - Libraries
  - Protocols

Input:
  - Target gene
  - Genome

Output:
  - Guides
  - Scores
""")

print("\n=== Cas-OFFinder ===")
print("""
Cas-OFFinder:
  - Command line
  - Fast
  - Genome-wide
  - Mismatches/bulges

Input:
  - Guide sequence
  - PAM
  - Genome
  - Mismatch threshold

Output:
  - Off-target sites
  - Mismatch count
  - Position

Command:
  cas-offinder guide.txt genome.fa out.txt
""")

print("\n=== Other Tools ===")
print("""
1. FlashFry
   - Fast
   - Genome-wide

2. GuideScan
   - Specificity
   - Genome-wide

3. CRISPRdirect
   - Simple
   - Web

4. E-CRISP
   - Web
   - Multiple genomes

5. GT-Scan
   - Web
   - Off-target
""")

print("\n=== Tool Comparison ===")
print(f"{'Tool':<20} {'Type':<15} {'Score':<15} {'Off-target':<15} {'Free':<10}")
print("-" * 75)
print(f"{'CRISPRscan':<20} {'Web':<15} {'Yes':<15} {'Yes':<15} {'Yes':<10}")
print(f"{'Benchling':<20} {'Web':<15} {'Yes':<15} {'Yes':<15} {'Academic':<10}")
print(f"{'CHOPCHOP':<20} {'Web':<15} {'Yes':<15} {'Yes':<15} {'Yes':<10}")
print(f"{'CRISPOR':<20} {'Web':<15} {'Yes':<15} {'Yes':<15} {'Yes':<10}")
print(f"{'Broad GPP':<20} {'Web':<15} {'Yes':<15} {'Yes':<15} {'Yes':<10}")
print(f"{'Cas-OFFinder':<20} {'CLI':<15} {'No':<15} {'Yes':<15} {'Yes':<10}")

print("\n=== When to Use Which ===")
print("""
USE CRISPRscan:
  ✓ Zebrafish
  ✓ In vivo
  ✓ Quick

USE Benchling:
  ✓ All-in-one
  ✓ Plasmid design
  ✓ Notebook

USE CHOPCHOP:
  ✓ Free
  ✓ Multiple genomes
  ✓ Quick

USE CRISPOR:
  ✓ Comprehensive
  ✓ Multiple scores
  ✓ Off-target

USE Broad GPP:
  ✓ Doench score
  ✓ Libraries
  ✓ CRISPRko/i/a

USE Cas-OFFinder:
  ✓ Off-target
  ✓ Genome-wide
  ✓ Command line
""")

🐍 Python Integration

🎯 CRISPR Design from Python

You can design guides, score them, and predict off-targets from Python. Here's how.

Python

Design: Find PAMs. Score: GC, structure. Off-target: Cas-OFFinder. Parse: Output. Visualize: Plot. Automate: Pipeline.

🐍 Design🐍 Score🐍 Off-target🐍 Automate

# Python Integration
print("=== Find PAMs ===")
print("""
def find_pams(sequence, pam="NGG"):
    \"\"\"Find PAM sites in a sequence.\"\"\"
    import re
    
    # Convert N to regex
    pattern = pam.replace('N', '[ATGC]')
    pattern = pattern.replace('R', '[AG]')
    pattern = pattern.replace('Y', '[CT]')
    pattern = pattern.replace('V', '[ACG]')
    
    guides = []
    for match in re.finditer(f'(?=({pattern}))', sequence):
        pam_start = match.start()
        pam_seq = match.group(1)
        
        # Extract 20nt upstream
        if pam_start >= 20:
            guide = sequence[pam_start-20:pam_start]
            guides.append({
                'guide': guide,
                'pam': pam_seq,
                'position': pam_start
            })
    
    return guides

# Example
seq = "ATGGATTTATCTGCTCTTCGCGTTGAAGAAGTACAAAATGTCATTAATGCTATGCAGAAAATCTTAGAGTGTCCCATCTGTCTGGAGTTGATCAAGGAACCTGTCTCCACAAAGTGTGACCACATATTTTGCAAATTTTGCATGCTGAAACTTCTCAACCAGAAGAAAGGGCCTTCACAGTGTCCTTTATGTAAGAATGATATAACCAAAAGGAGCCTACAAGAAAGTACGAGATTTAGTCAACTTGTTGAAGAGCTATTGAAAATCATTTGTGCTTTTCAGCTTGACACAGGTTTGGAGTATGCAAACAGCTATAATTTTGCAAAAAAGGAAAATAACTCTCCTGAACATCTAAAAGATGAAGTTTCTATCATCCAAAGTATGGGCTACAGAAACCGTGCCAAAAGACTTCTACAGAGTGAACCCGAAAATCCTTCCTTGCAGGAAACCAGTCTCAGTGTCCAACTCTCTAACCTTGGAACTGTGAGAACTCTGAGGACAAAGCAGCGGATACAACCTCAAAAGACGTCTGTCTACATTGAATTGGGATCTGATTCTTCTGAAGATACCGTTAATAAGGCAACTTATTGCAGTGTGGGAGATCAAGAATTGTTACAAATCA"
guides = find_pams(seq)
for g in guides[:5]:
    print(g)
""")

print("\n=== GC Score ===")
print("""
def gc_content(guide):
    \"\"\"Compute GC content.\"\"\"
    return (guide.count('G') + guide.count('C')) / len(guide)

def gc_score(guide):
    \"\"\"Score GC content (1 = optimal).\"\"\"
    gc = gc_content(guide)
    return 1 - abs(gc - 0.5) * 2

# Example
guides = [
    "ATGGATTTATCTGCTCTTCG",
    "TGGATTTATCTGCTCTTCGC",
    "GGATTTATCTGCTCTTCGCG",
]

for g in guides:
    print(f"{g}: GC={gc_content(g):.2f}, score={gc_score(g):.2f}")
""")

print("\n=== Filter Guides ===")
print("""
def filter_guides(guides, min_gc=0.4, max_gc=0.6, avoid="TTTT"):
    \"\"\"Filter guides by GC content and homopolymers.\"\"\"
    filtered = []
    for g in guides:
        gc = gc_content(g['guide'])
        if min_gc <= gc <= max_gc and avoid not in g['guide']:
            g['gc'] = gc
            filtered.append(g)
    return filtered

# Example
guides = find_pams(seq)
filtered = filter_guides(guides)
print(f"Total: {len(guides)}, Filtered: {len(filtered)}")
""")

print("\n=== Off-Target with Cas-OFFinder ===")
print("""
import subprocess

def run_cas_offinder(guide, genome, pam="NGG", mismatches=3):
    \"\"\"Run Cas-OFFinder for a guide.\"\"\"
    # Write input
    with open("guide.txt", "w") as f:
        f.write(f"{guide} {pam}\\n")
    
    # Run
    cmd = ['cas-offinder', 'guide.txt', genome, 'out.txt']
    subprocess.run(cmd, check=True)
    
    # Parse output
    off_targets = []
    with open("out.txt") as f:
        for line in f:
            parts = line.strip().split()
            if len(parts) >= 4:
                off_targets.append(parts)
    
    return off_targets

# Use
# off_targets = run_cas_offinder("ATGGATTTATCTGCTCTTCG", "genome.fa")
""")

print("\n=== Complete Pipeline ===")
print("""
def design_guides(sequence, pam="NGG", top_n=5):
    \"\"\"Complete guide design pipeline.\"\"\"
    # Find PAMs
    guides = find_pams(sequence, pam)
    
    # Filter
    filtered = filter_guides(guides)
    
    # Score
    for g in filtered:
        g['score'] = gc_score(g['guide'])
    
    # Sort
    filtered.sort(key=lambda x: x['score'], reverse=True)
    
    # Top N
    return filtered[:top_n]

# Use
guides = design_guides(seq)
for g in guides:
    print(f"{g['guide']} (GC={g['gc']:.2f}, score={g['score']:.2f})")
""")

print("\n=== Visualization ===")
print("""
import matplotlib.pyplot as plt

def plot_guides(guides):
    \"\"\"Plot guide scores.\"\"\"
    names = [g['guide'][:10] + '...' for g in guides]
    scores = [g['score'] for g in guides]
    
    plt.figure(figsize=(10, 5))
    plt.bar(names, scores, color='purple')
    plt.xlabel("Guide")
    plt.ylabel("Score")
    plt.title("Guide Scores")
    plt.xticks(rotation=45)
    plt.tight_layout()
    plt.savefig("guides.png")
    plt.show()

# Use
# plot_guides(guides)
""")

print("\n=== Order Oligos ===")
print("""
def order_oligos(guides):
    \"\"\"Generate oligo sequences for cloning.\"\"\"
    for i, g in enumerate(guides):
        guide = g['guide']
        # Add overhangs (example for lentiCRISPR)
        top = f"CACCG{guide}"
        bottom = f"AAAC{reverse_complement(guide)}C"
        print(f"Guide {i+1}:")
        print(f"  Top: {top}")
        print(f"  Bottom: {bottom}")

def reverse_complement(seq):
    \"\"\"Reverse complement.\"\"\"
    comp = {'A':'T', 'T':'A', 'G':'C', 'C':'G'}
    return ''.join(comp.get(b, 'N') for b in reversed(seq))

# Use
# order_oligos(guides)
""")

✂️ Editing Types (KO, KI, Base, Prime)

🎯 Types of CRISPR Editing

CRISPR can be used for various editing outcomes: knockout, knock-in, base editing, prime editing, and more. Each has specific guide design considerations.

Editing Types

KO: Frameshift. KI: HDR. Base: C→T, A→G. Prime: Precise. CRISPRi: Repress. CRISPRa: Activate.

✂️ Knockout✂️ Knock-in✂️ Base✂️ Prime

# Editing Types
print("=== 1. Knockout (KO) ===")
print("""
Goal: Disrupt gene function

Mechanism:
  - NHEJ repair
  - Frameshift
  - Premature stop

Guide design:
  - Early exon (5' end)
  - Functional domain
  - Conserved region

Example:
  Exon 1
  → Frameshift
  → Knockout

Tools:
  - CRISPRscan
  - Benchling
  - CHOPCHOP
""")

print("\n=== 2. Knock-in (KI) ===")
print("""
Goal: Insert sequence

Mechanism:
  - HDR repair
  - Donor template
  - Precise insertion

Guide design:
  - Near insertion site
  - HDR efficiency
  - Distance from cut

Example:
  Cut site
  → Donor template
  → Insertion

Tools:
  - Benchling
  - CRISPOR
  - HDR design
""")

print("\n=== 3. Base Editing ===")
print("""
Goal: Single base change

Mechanism:
  - dCas9 or nickase
  - Deaminase
  - C→T or A→G

Types:
  - CBE (C→T)
  - ABE (A→G)

Guide design:
  - Target window
  - PAM
  - No indels

Example:
  C→T
  → Point mutation

Tools:
  - Beam
  - Benchling
  - Base editing design
""")

print("\n=== 4. Prime Editing ===")
print("""
Goal: Precise editing

Mechanism:
  - Cas9 nickase
  - Reverse transcriptase
  - pegRNA

Guide design:
  - pegRNA
  - PBS
  - RTT

Example:
  pegRNA
  → Precise edit

Tools:
  - PrimeDesign
  - Benchling
  - pegRNA design
""")

print("\n=== 5. CRISPRi ===")
print("""
Goal: Repress transcription

Mechanism:
  - dCas9
  - KRAB
  - Repression

Guide design:
  - Promoter
  - TSS
  - Non-template strand

Example:
  dCas9-KRAB
  → Repression

Tools:
  - CRISPOR
  - Benchling
""")

print("\n=== 6. CRISPRa ===")
print("""
Goal: Activate transcription

Mechanism:
  - dCas9
  - Activator (VP64, p300)
  - Activation

Guide design:
  - Promoter
  - TSS
  - Template strand

Example:
  dCas9-VP64
  → Activation

Tools:
  - CRISPOR
  - Benchling
""")

print("\n=== 7. Multiplexing ===")
print("""
Goal: Multiple edits

Mechanism:
  - Multiple guides
  - Multiplex

Guide design:
  - Multiple targets
  - Compatible guides
  - Delivery

Example:
  3 guides
  → 3 edits

Tools:
  - Benchling
  - CRISPOR
""")

print("\n=== 8. Comparison ===")
print(f"{'Type':<15} {'Mechanism':<20} {'Guide':<20} {'Outcome':<15}")
print("-" * 70)
print(f"{'KO':<15} {'NHEJ':<20} {'Early exon':<20} {'Frameshift':<15}")
print(f"{'KI':<15} {'HDR':<20} {'Near site':<20} {'Insertion':<15}")
print(f"{'Base':<15} {'Deaminase':<20} {'Window':<20} {'Point mutation':<15}")
print(f"{'Prime':<15} {'RT':<20} {'pegRNA':<20} {'Precise':<15}")
print(f"{'CRISPRi':<15} {'dCas9-KRAB':<20} {'Promoter':<20} {'Repression':<15}")
print(f"{'CRISPRa':<15} {'dCas9-VP64':<20} {'Promoter':<20} {'Activation':<15}")

✅ Validation

🎯 Validating Your Edit

After designing and using a guide, you must validate that the edit worked and check for off-target effects.

Validation

On-target: T7, Surveyor, NGS. Off-target: GUIDE-seq, DISCOVER. Expression: qPCR, Western. Function: Phenotype. Karyotype: Rearrangements.

✅ On-target✅ Off-target✅ Expression✅ Function

# Validation
print("=== On-Target Validation ===")
print("""
Methods:
  1. T7 Endonuclease I
     - Cleaves mismatches
     - Gel electrophoresis
     - Simple, cheap

  2. Surveyor Assay
     - Similar to T7
     - Commercial kit

  3. Amplicon Sequencing
     - PCR + NGS
     - Quantify indels
     - Standard

  4. Whole-Genome Sequencing
     - Comprehensive
     - Expensive

  5. Sanger Sequencing
     - Clonal
     - Mixed population
""")

print("\n=== Off-Target Validation ===")
print("""
Methods:
  1. GUIDE-seq
     - Unbiased
     - Genome-wide
     - dsODN tag

  2. DISCOVER-seq
     - MRE11 binding
     - Genome-wide

  3. CIRCLE-seq
     - In vitro
     - Genome-wide

  4. SITE-seq
     - In vitro
     - Genome-wide

  5. Targeted sequencing
     - Predicted off-targets
     - Amplicon sequencing
""")

print("\n=== Expression Validation ===")
print("""
Methods:
  1. qRT-PCR
     - mRNA levels
     - Knockdown

  2. Western Blot
     - Protein levels
     - Knockout

  3. Flow Cytometry
     - Surface markers
     - Reporter

  4. Immunofluorescence
     - Localization
     - Expression
""")

print("\n=== Functional Validation ===")
print("""
Methods:
  1. Phenotype
     - Growth
     - Survival
     - Morphology

  2. Pathway analysis
     - Signaling
     - Metabolism

  3. Rescue
     - Re-express
     - Restore function

  4. Animal models
     - In vivo
     - Disease
""")

print("\n=== Karyotype Validation ===")
print("""
Methods:
  1. Karyotyping
     - Chromosomes
     - Rearrangements

  2. FISH
     - Fluorescence
     - Specific loci

  3. Optical mapping
     - Long-range
     - Structural variants
""")

print("\n=== Validation Workflow ===")
print("""
1. Design guide
2. Transfect cells
3. Harvest DNA
4. T7 assay (quick)
5. Amplicon sequencing (quantify)
6. Off-target (GUIDE-seq)
7. Expression (qPCR)
8. Function (phenotype)
9. Report
""")

print("\n=== Best Practices ===")
print("""
1. Multiple guides
   - 3-5 guides
   - Compare efficiency

2. Multiple methods
   - T7 + sequencing
   - Off-target + on-target

3. Controls
   - Non-targeting guide
   - Untransfected

4. Reproducibility
   - Biological replicates
   - Technical replicates

5. Report
   - Guide sequences
   - Scores
   - Validation
""")

⚖️ Tool Comparison

🎯 Which Tool?

Choosing the right guide design tool depends on organism, application, and resources.

Comparison

CRISPRscan: In vivo. Benchling: All-in-one. CHOPCHOP: Free. CRISPOR: Comprehensive. Broad GPP: Doench. Cas-OFFinder: Off-target.

⚖️ CRISPRscan⚖️ Benchling⚖️ CHOPCHOP⚖️ CRISPOR

Tool	Type	Score	Off-Target	Best For
CRISPRscan	Web	Yes	Yes	Zebrafish, in vivo
Benchling	Web	Yes	Yes	All-in-one
CHOPCHOP	Web	Yes	Yes	Free, multiple genomes
CRISPOR	Web	Yes	Yes	Comprehensive
Broad GPP	Web	Yes	Yes	Doench, libraries
Cas-OFFinder	CLI	No	Yes	Off-target
FlashFry	CLI	No	Yes	Fast, genome-wide
GuideScan	CLI	No	Yes	Specificity

# Decision Guide
print("=== Choose Tool by Organism ===")
print("""
Human:
  → Benchling
  → CRISPOR
  → Broad GPP

Mouse:
  → Benchling
  → CRISPOR
  → CRISPRscan

Zebrafish:
  → CRISPRscan
  → CHOPCHOP

Plant:
  → CHOPCHOP
  → CRISPR-P

Drosophila:
  → CHOPCHOP
  → FlyCRISPR

C. elegans:
  → CHOPCHOP
  → CRISPRscan
""")

print("\n=== Choose Tool by Application ===")
print("""
Knockout:
  → Benchling
  → CRISPRscan
  → CHOPCHOP

Knock-in:
  → Benchling
  → CRISPOR
  → HDR design

Base editing:
  → Beam
  → Benchling

Prime editing:
  → PrimeDesign
  → Benchling

Off-target:
  → Cas-OFFinder
  → CRISPOR
  → GUIDE-seq
""")

print("\n=== Speed Comparison ===")
print(f"{'Tool':<20} {'1 guide':<15} {'100 guides':<15} {'1000 guides':<15}")
print("-" * 65)
print(f"{'CRISPRscan':<20} {'< 1s':<15} {'< 10s':<15} {'< 1 min':<15}")
print(f"{'Benchling':<20} {'< 1s':<15} {'< 10s':<15} {'< 1 min':<15}")
print(f"{'CHOPCHOP':<20} {'< 1s':<15} {'< 10s':<15} {'< 1 min':<15}")
print(f"{'CRISPOR':<20} {'< 1s':<15} {'< 30s':<15} {'< 5 min':<15}")
print(f"{'Cas-OFFinder':<20} {'< 1s':<15} {'< 1 min':<15} {'< 10 min':<15}")

print("\n=== When to Use Which ===")
print("""
USE CRISPRscan:
  ✓ Zebrafish
  ✓ In vivo
  ✓ Quick

USE Benchling:
  ✓ All-in-one
  ✓ Plasmid design
  ✓ Notebook

USE CHOPCHOP:
  ✓ Free
  ✓ Multiple genomes
  ✓ Quick

USE CRISPOR:
  ✓ Comprehensive
  ✓ Multiple scores
  ✓ Off-target

USE Broad GPP:
  ✓ Doench score
  ✓ Libraries
  ✓ CRISPRko/i/a

USE Cas-OFFinder:
  ✓ Off-target
  ✓ Genome-wide
  ✓ Command line
""")

❓ FAQ

🎯 Frequently Asked Questions

Answers to common questions about CRISPR-Cas9 guide design.

What is a CRISPR-Cas9 guide RNA?

A CRISPR-Cas9 guide RNA (gRNA) is a short synthetic RNA (~20 nucleotides) that directs the Cas9 nuclease to a specific DNA sequence complementary to the guide. The gRNA forms a complex with Cas9, which then binds and cleaves the target DNA adjacent to a PAM sequence.

What is a PAM sequence?

A PAM (Protospacer Adjacent Motif) is a short DNA sequence (typically 5'-NGG-3' for SpCas9) that is required for Cas9 binding and cleavage. The PAM is located immediately 3' of the target sequence. Without a PAM, Cas9 cannot cleave the DNA.

How do I design a CRISPR guide RNA?

To design a gRNA: (1) Identify your target region, (2) Find PAM sequences (NGG for SpCas9), (3) Select a 20-nt sequence upstream of the PAM, (4) Check GC content (40-60%), (5) Avoid off-targets, (6) Check on-target score, (7) Avoid secondary structure, (8) Order oligos.

What makes a good CRISPR guide?

A good CRISPR guide has: (1) High on-target score, (2) Low off-target score, (3) GC content 40-60%, (4) No secondary structure, (5) No homopolymers (TTTT), (6) Target in early exons for knockout, (7) Good chromatin accessibility. Use tools like CRISPRscan, Benchling, or CHOPCHOP for scoring.

What is off-target effect in CRISPR?

Off-target effects occur when the guide RNA directs Cas9 to cleave unintended genomic sites that are similar (but not identical) to the target sequence. These can cause mutations elsewhere in the genome. Minimize off-targets by choosing guides with high specificity scores and verifying with tools like Cas-OFFinder.

How many guides should I design?

Design 3-5 guides per target. Not all guides work efficiently, so having multiple options increases your chances of success. Test them and choose the best one for your experiment.

What is the optimal GC content for a guide?

40-60% GC content is optimal. Too low (<40%) results in weak binding, while too high (>60%) can cause secondary structure and reduced efficiency.

Can I use CRISPR for base editing?

Yes. Base editors use a modified Cas9 (dCas9 or nickase) fused to a deaminase. They can convert C→T (CBE) or A→G (ABE) without creating double-strand breaks. Guide design is similar but with a target window.

What is prime editing?

Prime editing uses a Cas9 nickase fused to reverse transcriptase and a prime editing guide RNA (pegRNA). It allows precise insertions, deletions, and substitutions without double-strand breaks. Guide design is more complex (pegRNA with PBS and RTT).

How do I validate my CRISPR edit?

Validate with: T7 endonuclease assay (quick), amplicon sequencing (quantify), GUIDE-seq (off-target), qPCR/Western (expression), and phenotype (function). Always use multiple methods.

# Quick Reference
print("=== Quick Answers ===")
print("""
Q: What is gRNA?
A: RNA that guides Cas9 to target.

Q: What is PAM?
A: NGG for SpCas9; required for cleavage.

Q: How to design?
A: Find PAM, extract 20nt, check GC, off-target.

Q: Good guide?
A: High on-target, low off-target, GC 40-60%.

Q: Off-target?
A: Unintended cleavage; minimize with tools.

Q: How many guides?
A: 3-5 per target.

Q: Optimal GC?
A: 40-60%.

Q: Base editing?
A: C→T or A→G; dCas9 + deaminase.

Q: Prime editing?
A: Precise; pegRNA + RT.

Q: Validate?
A: T7, sequencing, GUIDE-seq, qPCR.

Q: Free tools?
A: Yes — CRISPRscan, CHOPCHOP, CRISPOR.
""")

🚀 Future

🎯 Where CRISPR Guide Design Is Going

The future is AI-designed guides, base and prime editing, in vivo delivery, and clinical applications.

Future Directions

AI: Learned design. Base: Precise. Prime: Versatile. In vivo: Delivery. Clinical: Therapies. Multiplex: Many edits.

🚀 AI🚀 Base🚀 Prime🚀 Clinical

# Future of CRISPR
print("=== 1. AI-Designed Guides ===")
print("""
Trend: Deep learning

Ideas:
  - Learned scoring
  - Neural networks
  - Transformer models
  - Protein language models

Examples:
  - CRISPR-GPT
  - DeepCRISPR
  - AI-designed guides

Future:
  - Better efficiency
  - Better specificity
  - Cell-type specific
  - Context-aware
""")

print("\n=== 2. Base Editing ===")
print("""
Trend: Precise base changes

Current:
  - CBE (C→T)
  - ABE (A→G)

Future:
  - C→G
  - A→T
  - Transversion

Examples:
  - BE4max
  - ABE8e
  - New editors

Future:
  - All transitions
  - All transversions
  - Precise editing
""")

print("\n=== 3. Prime Editing ===")
print("""
Trend: Precise editing

Current:
  - Insertions
  - Deletions
  - Substitutions

Future:
  - Large insertions
  - Complex edits
  - Multiplex

Examples:
  - PE2, PE3
  - PrimeDesign
  - pegRNA optimization

Future:
  - Higher efficiency
  - Larger edits
  - Clinical
""")

print("\n=== 4. In Vivo Delivery ===")
print("""
Trend: Delivery

Current:
  - Ex vivo
  - AAV
  - LNP

Future:
  - Tissue-specific
  - Organ-targeted
  - Non-viral

Examples:
  - Lipid nanoparticles
  - AAV vectors
  - Virus-like particles

Future:
  - In vivo editing
  - Tissue-specific
  - Clinical
""")

print("\n=== 5. Clinical Applications ===")
print("""
Trend: Therapies

Current:
  - Casgevy (sickle cell)
  - Clinical trials

Future:
  - Cancer
  - Genetic diseases
  - Infectious diseases

Examples:
  - Sickle cell
  - Beta-thalassemia
  - Leber congenital amaurosis
  - Huntington's

Future:
  - More approvals
  - In vivo
  - Cost reduction
""")

print("\n=== 6. Multiplex Editing ===")
print("""
Trend: Many edits

Current:
  - 1-3 guides

Future:
  - 10+
  - Arrays
  - Combinatorial

Examples:
  - Metabolic engineering
  - Synthetic biology
  - Cell therapy

Future:
  - Complex edits
  - Pathways
  - Genomes
""")

# TIMELINE
print("\n=== Timeline ===")
print("""
Past:
  2012: CRISPR-Cas9
  2013: Mammalian editing
  2020: Nobel Prize
  2023: First therapy approved

Present:
  2025: Base/prime editing
  2026: AI-designed guides

Future:
  2030: In vivo therapies
  2035: Multiplex
  2040: Genome synthesis?
""")

print("\n=== Final Thoughts ===")
print("""
CRISPR guide design is:
  - Critical for success
  - On-target efficiency
  - Off-target safety
  - PAM-dependent
  - Evolving rapidly

The future is:
  - AI-designed
  - Base editing
  - Prime editing
  - In vivo
  - Clinical

"Design well, edit precisely."

Keep exploring. Keep building.
""")