Cuttingenes
Live · 3 enabled nucleases · Reproducible validation

Design CRISPR guides with confidence

CRISPR guide design platform. SpCas9, SaCas9 and Cas12a are commercially available, with paired-site off-target analysis for SpCas9. Cas13d is implemented but disabled until its model licence is resolved.

🏢 In-house deployment availableNeed privacy or regulatory compliance? CuttinGenes can be deployed in-house, on your own servers or private cloud, with no data exposure to third parties. Contact us about institutional in-house licenses. Coming soon: hardware appliance (plug & play, no IT required) for pharma and hospital environments.
3
nucleases enabled
r=0.8499
Pearson SpCas9 DeepHF
hg38
Off-target with real coordinates
11
analysis modules

New in the app

Lab protocol generator: from validated guide to oligo order in one click

Includes pX330/pX458 cloning oligos, PCR/Sanger primers, and expected amplicon size.

Try protocol

3 commercially enabled nucleases

SpCas9, SaCas9 and Cas12a are available; Cas13d remains a research implementation pending a commercial licence. For Cas12a the score is a relative-activity percentile among guides for the same target, not an editing percentage.

✂️

SpCas9

PAM: NGG

r = 0.8499
Pearson (DeepHF hold-out)
✂️

SaCas9

PAM: NNGRRT

r = 0.562
Pearson (Najm 2018, 293T)
✂️

Cas12a

PAM: TTTV

51 %
best-of-3 accuracy (chance 33 %)

Everything you need in one tool

From sequence to optimal experimental design in under one second.

Efficiency prediction

SpCas9 XGBoost trained on DeepHF (Wang 2019, CC BY; over 55,000 guides). Pearson r = 0.8499 on its own hold-out.

🎯

Off-target risk

SpCas9 guide-site pair scoring (CFD, seed, mismatches, PAM) plus real genome-wide search with Cas-OFFinder. It does not replace experimental validation.

🔬

Multi-nuclease

SpCas9, SaCas9 and Cas12a are available, all three trained on commercially licensed data. Cas13d remains disabled while its licence is negotiated.

✏️

Base editing (CBE/ABE)

Predict efficiency and bystander risk for C→T and A→G base editors. Position-by-position activity window analysis.

🧬

Gene → Guides

Type BRCA1 or TP53 and get the best guides instantly from Ensembl or NCBI — no FASTA download required.

🔗

REST API + Stripe billing

Integrate via API keys. Upgrade to Basic or Advanced directly from the app with secure Stripe payments.

11 integrated analysis modules

Everything in one interface. No need to switch tools.

🧬

Sequence prediction

Analyse one or more guides (20 nt): efficiency, off-target, CrisprScore and molecular features.

📊

Batch prediction (CSV)

Upload a CSV with hundreds of guides and get predictions for all of them in seconds.

🔬

PAM / FASTA scan

Upload a FASTA and get candidate guides with on-target efficiency; off-target scoring is available for SpCas9.

⚖️

Compare guides

Visualise and compare multiple guides side by side with interactive charts.

📋

Format CSV

Clean and normalise CSV files from any format to prepare them for prediction.

Experimental validation

Compare predictions against your own experimental data and calculate performance metrics (Pearson, Spearman, AUC).

🕐

Prediction history

Access all your previous predictions and download them at any time.

🔬

Multi-nuclease scan

Scan a FASTA with SpCas9, SaCas9 or Cas12a and filter by on-target activity. Cas13d will be enabled once its licence is resolved.

✏️

Base editing (CBE / ABE)

Predict efficiency and bystander risk for C→T and A→G editors with position-by-position window analysis.

🧬

Gene → Guides auto

Type BRCA1 or TP53 and get the best guides from Ensembl or NCBI without downloading any file.

🧪

Lab protocol generator

Generate cloning oligos (pX330/pX458), PCR/Sanger primers, and expected amplicon size to move from design to bench in one click.

🤖 Built-in AI assistant

Say it in your own words and the form fills itself

A conversational assistant inside the app. Describe what you need and it opens the right tab with the fields already set. It understands English and Spanish.

🔒

Your data never leaves the server

No external APIs, no third-party calls: all processing happens on the installation running the application — including the language model on deployments that ship it.

It proposes, it doesn't execute

It sets the form up and stops there. You review the values and press the button — nothing it says ever launches an analysis or spends quota.

📎

It takes your files

Drop a FASTA or a CSV into the chat and it places it in the right tab. The file stays in your browser until you run it.

🧬

It never invents sequences

Any guide in a field is one you typed. Ask for an example and it uses published guides, telling you that's what they are.

🤖Built-in AI assistant
guides for BRCA1 in human with SaCas9
I opened Gene → Guides and filled in Nuclease = SaCas9, Species = human, Gene name = BRCA1.
🏆 Externally validated against published tools

Performance validated on independent datasets

SpCas9 evaluated on data never seen during training. Competitive with state-of-the-art published tools.

#2 of 6
Labuhn 2018 ranking
Vs Doench2016, CHOPCHOP, CRISPRscan, sgRNA Designer and FlashFry
r = 0.276
External validation Labuhn
Pearson on 476 guides from the Labuhn 2018 dataset (unseen during training)
r = 0.275
External validation Xu 2015
Pearson on 2,069 guides from the Xu 2015 dataset (unseen during training)

How it works

01

Choose your input

Paste a 20 nt sequence, upload a CSV, type a gene name (BRCA1, TP53) or upload a FASTA. The tool adapts to your workflow.

02

Choose nuclease and editor

SpCas9 by default. Switch to SaCas9 or Cas12a. Enable base editing (CBE/ABE) for single-nucleotide changes without DNA breaks.

03

Get results and export

Review activity, position and strand; compatible SpCas9 flows also include off-target scoring and CrisprScore. Sort and export to CSV.

💡 Looking for something you don't see?

Tell us what you need from a CRISPR tool

If your team needs a feature, nuclease or integration that isn't here, describe your case. We'll review it and reach out to understand it better. It's not a feature promise: it's the start of a conversation.

Start for free today

The Free plan includes 50 predictions per month, no credit card required. Scale up with Basic (€49/mo) or Advanced (€149/team).