GEMINI: a variational Bayesian approach to identify genetic interactions from combinatorial CRISPR screens

Area of Science:

• Genetics
• Computational Biology
• Bioinformatics

Background:

• CRISPR-based combinatorial perturbation is a powerful tool for uncovering genetic interactions.
• Systematic identification of these relationships is challenging due to variability in samples, reagents, and biological processes.

Purpose of the Study:

• To develop a robust computational method for analyzing pairwise knockout screens.
• To improve the accuracy and scalability of identifying genetic interactions from CRISPR screens.

Main Methods:

• Developed a variational Bayes approach named GEMINI.
• GEMINI jointly analyzes all samples and reagents to identify genetic interactions.
• Implemented GEMINI as an open-source R package.

Main Results:

• GEMINI enhances the accuracy and scalability of genetic interaction identification.
• The method enables systematic analysis of combinatorial CRISPR knockout screens.
• GEMINI is adaptable to various screen designs and dimensions.

Conclusions:

• GEMINI provides a powerful and versatile tool for analyzing CRISPR-based genetic interaction screens.
• The open-source availability facilitates broader adoption and research in the field.
• This approach overcomes key challenges in systematic genetic interaction mapping.