Bioinformatics (Oxford, England)Nov 14, 2018

Using deep learning to predict unintended gene changes in CRISPR-Cas9 editing

CRISPR Gene Editing Weekly Brief ↗PubMed ↗DOI ↗OA ↗

Updated Mar 28, 2026

Abstract

The convolutional neural network model achieved an average classification area under the ROC curve (AUC) of 97.2%.

Two algorithms using deep neural networks were designed to predict off-target mutations in CRISPR-Cas9 gene editing.
The models were evaluated on the CRISPOR dataset and another dataset from GUIDE-seq.
The deep feedforward neural network also demonstrated competitive performance with an average AUC of 97.0%.
The proposed algorithms outperformed existing state-of-the-art off-target prediction methods and traditional machine learning models in AUC values.
Additional analyses were performed to explore the underlying reasons for the performance differences between the models.

Simplified

MOTIVATION: The prediction of off-target mutations in CRISPR-Cas9 is a hot topic due to its relevance to gene editing research. Existing prediction methods have been developed; however, most of them just calculated scores based on mismatches to the guide sequence in CRISPR-Cas9. Therefore, the existing prediction methods are unable to scale and improve their performance with the rapid expansion of experimental data in CRISPR-Cas9. Moreover, the existing methods still cannot satisfy enough precision in off-target predictions for gene editing at the clinical level.

RESULTS: To address it, we design and implement two algorithms using deep neural networks to predict off-target mutations in CRISPR-Cas9 gene editing (i.e. deep convolutional neural network and deep feedforward neural network). The models were trained and tested on the recently released off-target dataset, CRISPOR dataset, for performance benchmark. Another off-target dataset identified by GUIDE-seq was adopted for additional evaluation. We demonstrate that convolutional neural network achieves the best performance on CRISPOR dataset, yielding an average classification area under the ROC curve (AUC) of 97.2% under stratified 5-fold cross-validation. Interestingly, the deep feedforward neural network can also be competitive at the average AUC of 97.0% under the same setting. We compare the two deep neural network models with the state-of-the-art off-target prediction methods (i.e. CFD, MIT, CROP-IT, and CCTop) and three traditional machine learning models (i.e. random forest, gradient boosting trees, and logistic regression) on both datasets in terms of AUC values, demonstrating the competitive edges of the proposed algorithms. Additional analyses are conducted to investigate the underlying reasons from different perspectives.

AVAILABILITY AND IMPLEMENTATION: The example code are available at https://github.com/MichaelLinn/off_target_prediction. The related datasets are available at https://github.com/MichaelLinn/off_target_prediction/tree/master/data.

Key numbers

97.2%

CNN AUC

Average AUC on the CRISPOR dataset.

97.0%

FNN AUC

Average AUC on the CRISPOR dataset.

5.8%

Performance Improvement

AUC increase compared to the best traditional method.

Full Text

We can’t show the full text here under this license.

View Full Text ↗

Using deep learning to predict unintended gene changes in CRISPR-Cas9 editing

Abstract

Key numbers

Full Text

You found one interesting study. We’ll send the next 7.

what lands in your inbox each week:

Recent issues from the crispr gene editing brief

Abstract

Key numbers

Full Text

Related papers

You found one interesting study. We’ll send the next 7.

what lands in your inbox each week:

Recent issues from the crispr gene editing brief