EpiCas-DL: Predicting sgRNA activity for CRISPR-mediated epigenome editing by deep learning

Area of Science:

• Molecular Biology
• Bioinformatics
• Gene Therapy

Background:

• CRISPR epigenome editing offers precise gene expression control without DNA modification, crucial for research and therapeutics.
• Efficient design of single guide RNA (sgRNA) is vital for maximizing the efficacy and specificity of epigenome editing tools.

Purpose of the Study:

• To develop and validate a deep learning framework, EpiCas-DL, for optimizing sgRNA design in CRISPR-mediated epigenome editing.
• To enhance the prediction accuracy of sgRNA activity for both gene silencing and gene activation.

Main Methods:

• Extensive analysis of CRISPR epigenome editing tools across thousands of experimentally validated on-target sites.
• Development of EpiCas-DL, a deep learning framework utilizing sequence and epigenetic features for sgRNA design optimization.

Main Results:

• EpiCas-DL demonstrates high accuracy in predicting sgRNA activity for gene silencing and activation.
• The framework outperforms existing in silico methods for sgRNA design.
• Identification of key epigenetic and sequence features influencing sgRNA efficacy.

Conclusions:

• EpiCas-DL provides a powerful and accurate tool for designing sgRNAs for CRISPR epigenome editing.
• The findings facilitate the broader application of epigenome editing in basic research and therapeutic strategies.
• The EpiCas-DL framework is publicly accessible for researchers.