Prediction of genome base-editing efficiency and outcomes based on machine learning: A deep review

Base editing, a revolutionary genome editing technology, has risen to prominence for its distinguished features such as high fidelity, precision, and targeted specificity. It has found broad applications across the spectrum of gene therapy, precise breeding, and in-depth gene function studies. The e...

Full description

Saved in:
Bibliographic Details
Published in:Journal of biotech research 2024-01, Vol.18, p.183-201
Main Authors: Yao, Haichang, Hu, Guangyong, Zhang, Peihao
Format: Article
Language:English
Subjects:
CRISPR
Deep learning
Editing
Editors
Efficiency
Enzymes
Gene therapy
Genome editing
Learning algorithms
Machine learning
Mutation
Neural networks
Predictions
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Base editing, a revolutionary genome editing technology, has risen to prominence for its distinguished features such as high fidelity, precision, and targeted specificity. It has found broad applications across the spectrum of gene therapy, precise breeding, and in-depth gene function studies. The efficiency of base editing and the integrity of resultant genotypic products are the most important performances of base editing technology, which determine whether it can ultimately be suitable for clinical utilization. Because the underlying determinants that influence base editing efficiency and genotypic output remain elusive, the optimization of base editing presently is predominantly dependent on empirical knowledge and iterative experimental attempts. Machine-learning-based prediction for editing efficiency and genotypic outputs can guide base editing applications and optimize base editors in silico, helping researchers improving experimental efficiency and saving experimental costs, which positions it as a significant research direction within this field. This research systematically reviewed the development trajectory of prediction methodologies from CRISPR/Cas9 to base editing, highlighted the intrinsic differences between predictions for base editing and those for CRISPR/Cas9, and then provided a detailed review of all outstanding base editing prediction methods for the first time. The key issues and future directions were also provided for upcoming researchers.
ISSN:1944-3285