TABLE 3.
Summary of the Contributions of Deep Learning Studies in Predicting gRNA Activities.
| Author | Contributions |
|---|---|
| Ameen et al. (2021) | Proposed a hybrid CNN-SVR model for predicting gRNA activity in the CRISPR-Cas12 system, outperforming existing models |
| Shrawgi and Sisodia. (2019) | Developed a CNN model, DeepSgRNA, to predict the efficiency of sgRNAs in the CRISPR-Cas9 system |
| Dimauro et al. (2019) | Introduced CRISPRLearner, a CNN-based model for predicting sgRNA cleavage efficiency |
| Xie et al. (2023) | Proposed CRISPR-OTE, a CNN and biLSTM-based framework for gRNA on-target efficiency prediction |
| Xue et al. (2019) | Presented DeepCas9, a CNN-based model for identifying functional sgRNAs in the CRISPR-Cas9 system |
| Wan and Jiang. (2023) | Developed TransCrispr, a Transformer and CNN-based model for predicting sgRNA knockout efficacy |
| Zhang et al. (2020b) | Proposed a hybrid CNN-SVR system for predicting gRNA on-target efficacy in the CRISPR-Cas9 system |
| Li et al. (2022a) | Introduced a CNN and XGBoost-based model, CNN-XG, for predicting sgRNA on-target knockout efficacy |
| Xiang et al. (2021) | Created CRISPRon, a deep learning model for advanced gRNA efficiency predictions |
| Kirillov et al. (2022) | Presented a hybrid of Capsule Networks and Gaussian Processes for predicting gRNA cleavage efficiency |
| Elkayam and Orenstein. (2022) | Developed DeepCRISTL, a transfer learning model for predicting on-target editing efficiency in the CRISPR-Cas9 system |
| Niu et al. (2021b) | Introduced R-CRISPR, a deep learning model for predicting off-target activities in CRISPR-Cas9 |
| Luo et al. (2019) | Presented DeepCpf1, a deep CNN model for predicting CRISPR-Cpf1 gRNAs on-target activity and off-target effects |
| Zhang et al. (2021a) | Proposed interpretable attention-based CNN models, CRISPR-ONT and CRISPR-OFFT, for predicting CRISPR-Cas9 sgRNA activities |
| Yang et al. (2023) | Developed EpiCas-DL, a deep learning framework for optimizing sgRNA design for CRISPR-mediated epigenome editing |
| Zhang et al. (2020c) | Developed DL-CRISPR, a deep learning model for predicting off-target activity in CRISPR-Cas9 |
| Liu et al. (2020) | Introduced CnnCrispr, a model for predicting off-target propensity of sgRNA. |
| Zhang and Jiang. (2022) | Introduced CRISPR-IP, a CNN, BiLSTM, and attention layers-based model for CRISPR-Cas9 |
| Charlier et al. (2021) | Introduced a novel encoding of sgRNA-DNA sequences to enhance deep learning off-target prediction in CRISPR-Cas9 gene editing |
| Lin et al. (2022) | Presented an AI approach integrating CNNs and attention module for quantifying CRISPR gene-editing off-target effects |
| Niu et al. (2021a) | Proposed an ensemble CNN model, sgRNACNN, for identifying high on-target activity of sgRNA in four agronomic species |
| Jost et al. (2020) | Used CRISPR interference to control gene expression, deriving rules governing mismatched sgRNA activity using deep learning |
| Xiao et al. (2021) | Proposed AttCRISPR, an interpretable model to predict sgRNA on-target activity, integrating encoding and embedding-based methods |
| Wang and Zhang. (2019) | Designed a CNN for predicting sgRNA activity in E. coli, emphasizing the importance of species-specific models in sgRNA prediction |