CasMiner: a deep-learning tool for high-throughput mining and rational design of efficient Cas9

Area of Science:

• Biotechnology
• Genomics
• Molecular Biology

Background:

• CRISPR-Cas9 technology is vital for life sciences, but discovering new Cas9 variants is challenging.
• Current methods for Cas9 discovery and optimization rely on sequence alignment, limiting innovation.

Purpose of the Study:

• To develop a deep-learning model, CasMiner, for identifying and engineering novel Cas9 proteins.
• To validate the efficacy of newly discovered and engineered Cas9 variants for genome editing.

Main Methods:

• Developed CasMiner, a deep-learning model for Cas9 protein prediction and engineering.
• Identified VpCas9 from databases and performed experimental validation of its cleavage activity.
• Engineered VpCas9 mutants using CasMiner and evolutionary analysis, focusing on structural rigidity and charge.
• Conducted in vivo cleavage assays in rice and maize to evaluate editing efficiency.

Main Results:

• CasMiner achieved 99.63% accuracy in predicting Cas9 proteins.
• VpCas9 demonstrated robust double-strand DNA cleavage activity.
• Engineered VpCas9 mutant VPM2-3 exhibited superior editing efficiency in rice and maize compared to wild-type VpCas9.
• The editing efficiency of VPM2-3 rivaled that of the standard SpCas9.

Conclusions:

• CasMiner provides a comprehensive platform for discovering and engineering Cas9 proteins.
• VpCas9 and its engineered mutants offer powerful new tools for advancing genome editing applications.
• This work significantly expands the toolkit available for precise and efficient genome modification.