CASowary: CRISPR-Cas13 guide RNA predictor for transcript depletion.

Area of Science:

• Molecular Biology
• Bioinformatics

Background:

• CRISPR-Cas systems, including Cas13, enable precise RNA editing.
• Predicting single guide RNA (sgRNA) efficacy for RNA knockdown is challenging due to transcriptome complexity and RNA structures.
• Understanding factors influencing sgRNA effectiveness is crucial for transcript targeting.

Purpose of the Study:

• To develop a computational tool, CASowary, for predicting sgRNA efficacy in Cas13-mediated RNA knockdown.
• To identify sequence and target availability features that determine sgRNA performance.

Main Methods:

• Utilized a Decision Tree machine learning model trained on 555 sgRNAs and RNA knockdown data from Cas13 experiments in HEK293 cells.
• Incorporated transcriptome-wide protein occupancy data.
• Features included 112 sequence and target availability parameters.

Main Results:

• CASowary achieved a noise-normalized accuracy exceeding 70% in classifying sgRNA efficacy into four knockdown levels.
• High-efficacy sgRNA predictions were validated using the CIRTS RNA targeting system.
• The tool demonstrated cell-line-specific prediction of high-quality guides by integrating protein occupancy data.

Conclusions:

• CASowary facilitates the rapid selection of effective sgRNAs for CRISPR/Cas13 applications.
• Enables development of RNA-targeting therapeutics for diseases involving RNA regulatory aberrations.