Table 1.
CRISPR and Machine Learning/Deep Learning related reviews
| Article | Year range | Papers | Scope | Shortcomings |
|---|---|---|---|---|
| [41] | 2017–2022 | 57 | This study focuses on ML techniques to predict CRISPR/Cas9 sgRNA activity (on/off-target cleavage), to assist sgRNA design and identify current research trends. | The study is limited to a systematic mapping, excluding comparisons of methods or results. |
| [99] | 2019–2023 | 54 | This review article focuses on the applications of DL in multiple aspects of CRISPR-Cas, the prime focus is on gRNA activity prediction, CRISPR-Cas editing outcomes, design of High-Activity gRNAs, Automated System Implementation, Nucleic Acid Detection, Anti-CRISPR Protein Identification, Cas9 Variant Activity Prediction, Transcription Factor Binding Prediction | Not all topics are equally focused on. ML models, feature representation methods, and publicly available CRISPR-Cas associated benchmark datasets are not discussed. |
| [76] | 2016–2019 | 11 | Future of CRISPR-based biosensors, genome engineering, discovery of CRISPR, conventional biosensors, IoT, Big Biomedical Data, Cloud Computing Systems, integration of AI in CRISPR-based biosensors | There is no discussion on the use of AI in CRISPR. |
| [13] | – | – | Applicability of CRISPR/Cas9 in cancer research, CRISPR/Cas9 in drug resistance, CRISPR clinical trials, on/off-target gRNA activity prediction | The focus in biological/biochemical aspects is much bigger than on AI |
| [151] | till 2022 | – | ML models in cancer, limited CRISPR details, drug discovery through AI/ML, precision and genomic medicine, different ML Models | Deep Learning is not described in detail and there is only a small dicussion of CRISPR |
| [177] | 2014–2022 | 15 | CRISPR for breast cancer treatment, AI/ML for therapy strategy, on/off-target effects of gRNA | Specific focus on Triple Negative Breast Cancer, no other fields than on/off target effects are dealt with |
| [84] | 2017–2022 | 21 | A perspective on AI in CRISPR/Cas9 modification, gRNA design, clinical trials. It explores how AI can enhance CRISPR’s precision and effectiveness in treating genetic diseases, particularly cancer, while also examining the current limitations and future possibilities of this approach. | This perspective study does not discuss any details of benchmark datasets, feature engineering approaches, and ML or DL methods. |
| [144] | – | - | ML effects on CRISPR gene editing, data labeling pitfalls, data selection, feature engineering, gRNA design and effects prediction | Only on/off-target activity prediction is discussed |
| [164] | 2014–2022 | 49 | ML/DL models in CRISPR/Cas9, on/off target activity prediction, data preprocessing, gRNA encoding | Only on/off-target activity prediction is discussed |
| [64] | 2017–2021 | 9 | AI in designing gene delivery vehicles, improving CRISPR/Cas, nanobots and mRNA vaccine carriers develpment | No other fields than on/off target effects are dealt with |
| [93] | 2015–2021 | 20 | On-target activity prediction, gRNA design, DL tools evaluation, comparison of learning based (DL) and hypothesis driven tools | No other fields than on target effects are dealt with |
| [194] | till 2019 | 20 | ML/DL algorithms for on/off target prediction, gRNA design, challenges in CRISPR activity and specificity prediction | No other fields than on target effects are dealt with |
| [162] | till 2023 | – | ML and DL models in PAM prediction, gRNA designing, on/off-target activity prediction, and Prime editing and pegRNA designing | Important details related to datasets, representation learning, and ML and DL models are missing. In addition, only 5 different of AI in CRISPR are covered. |