Related Experiment Video
Updated: Aug 10, 2025

CIRCLE-Seq for Interrogation of Off-Target Gene Editing
Published on: November 1, 2024
Modeling CRISPR-Cas13d on-target and off-target effects using machine learning approaches
Xiaolong Cheng1,2, Zexu Li3, Ruocheng Shan1,4
1Center for Genetic Medicine Research, Children's National Hospital, Washington, DC, 20010, USA.
Abstract:
A major challenge in the application of the CRISPR-Cas13d system is to accurately predict its guide-dependent on-target and off-target effect. Here, we perform CRISPR-Cas13d proliferation screens and design a deep learning model, named DeepCas13, to predict the on-target activity from guide sequences and secondary structures. DeepCas13 outperforms existing methods to predict the efficiency of guides targeting both protein-coding and non-coding RNAs. Guides targeting non-essential genes display off-target viability effects, which are closely related to their on-target efficiencies. Choosing proper negative control guides during normalization mitigates the associated false positives in proliferation screens. We apply DeepCas13 to the guides targeting lncRNAs, and identify lncRNAs that affect cell viability and proliferation in multiple cell lines. The higher prediction accuracy of DeepCas13 over existing methods is extensively confirmed via a secondary CRISPR-Cas13d screen and quantitative RT-PCR experiments. DeepCas13 is freely accessible via http://deepcas13.weililab.org .
Insights
We developed DeepCas13, a deep learning model to predict CRISPR-Cas13d guide efficiency for RNA targeting. This tool improves accuracy in predicting on-target and off-target effects, aiding gene function studies.
Area of Science:
- Molecular Biology
- Genomics
- Bioinformatics
Background:
- CRISPR-Cas13d systems offer powerful RNA manipulation tools.
- Accurate prediction of guide RNA on-target and off-target effects remains a significant challenge.
Purpose of the Study:
- To develop a deep learning model, DeepCas13, for predicting CRISPR-Cas13d on-target activity.
- To enhance the prediction of guide efficiency for targeting both protein-coding and non-coding RNAs.
Main Methods:
- CRISPR-Cas13d proliferation screens were conducted.
- A deep learning model, DeepCas13, was designed using guide sequences and RNA secondary structures.
- Model performance was validated using secondary screens and quantitative RT-PCR.
Main Results:
- DeepCas13 demonstrated superior performance over existing methods in predicting guide efficiency.
- Off-target viability effects were observed for guides targeting non-essential genes, correlating with on-target efficiency.
- lncRNAs impacting cell viability and proliferation were identified using DeepCas13.
Conclusions:
- DeepCas13 accurately predicts CRISPR-Cas13d guide activity, improving RNA targeting predictions.
- Proper normalization strategies using negative controls are crucial for mitigating false positives in proliferation screens.
- The study identifies novel lncRNAs involved in cell viability and proliferation.
Related Concept Videos
CRISPR
CRISPR/Cas9 Genome Editing
CRISPR and crRNAs
The CRISPR-Cas system stores a copy of foreign DNA in the host genome and uses it to identify the foreign DNA upon reinfection. CRISPR-Cas has three different...

