Related Experiment Video
Updated: Sep 11, 2025
07:31
Efficient PAM-Less Base Editing for Zebrafish Modeling of Human Genetic Disease with zSpRY-ABE8e
Published on: February 17, 2023
1.3K
Systematic pegRNA design with PRIDICT2.0 and ePRIDICT for efficient prime editing
Nicolas Mathis1, Kim Fabiano Marquart2, Ahmed Allam3
1Institute of Pharmacology and Toxicology, University of Zurich, Zurich, Switzerland. nicolas.mathis@pharma.uzh.ch.
Nature Protocols
|August 15, 2025
Summary
Computational tools PRIDICT2.0 and ePRIDICT predict prime editing guide RNA (pegRNA) efficiency, streamlining genetic modifications. These machine-learning models optimize prime editing by analyzing pegRNA design and chromatin context, accelerating research.
Area of Science:
- Molecular Biology
- Genomics
- Bioinformatics
Background:
- Prime editing is a precise genome editing technology that avoids DNA double-strand breaks.
- Optimizing prime editing guide RNA (pegRNA) design is crucial but experimentally challenging due to numerous variables.
- Existing computational tools have limitations in accommodating large edits or considering chromatin context.
Purpose of the Study:
- To develop and present computational tools, PRIDICT2.0 and ePRIDICT, for predicting prime editing efficiency.
- To streamline the selection of optimal pegRNAs by assessing pegRNA design and chromatin context.
- To facilitate the adoption of prime editing in basic and translational research.
Main Methods:
- PRIDICT2.0: An ensemble of attention-based bidirectional recurrent neural networks predicting pegRNA efficiencies for various edit types (replacements, insertions, deletions) up to 40 base pairs.
- ePRIDICT: A gradient-boosting algorithm incorporating local chromatin environments and genomic location effects on prime editing rates.
- Both tools are accessible via a web server (www.pridict.it) for individual predictions and available for local installation for batch processing.
Main Results:
- PRIDICT2.0 accurately predicts pegRNA efficiencies, accommodating larger edits and enabling silent bystander edits to enhance efficiency.
- ePRIDICT integrates chromatin context, providing a more comprehensive assessment of editing potential at specific genomic loci.
- Web-based predictions are rapid (under a minute), while local batch processing offers scalability for extensive analyses.
Conclusions:
- PRIDICT2.0 and ePRIDICT significantly simplify and accelerate the process of pegRNA design and selection for prime editing.
- These machine-learning tools enhance the efficiency and applicability of prime editing by considering both sequence and epigenetic factors.
- The developed computational resources are expected to boost the utilization of prime editing technology in diverse research settings.
