Discovering functional sequences with RELICS, an analysis method for CRISPR screens.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• CRISPR screens identify genome sequences impacting cellular phenotypes like gene expression and survival.
• Analyzing non-coding sequences with CRISPR screens offers potential for discovering novel functional elements.
• Current analysis tools are insufficient for the complexities of non-coding CRISPR screens.

Purpose of the Study:

• To introduce RELICS, a Bayesian hierarchical model designed for discovering functional sequences from CRISPR screens.
• To address challenges in non-coding CRISPR screens, including unknown functional sequence locations and data overdispersion.
• To improve the precision, recall, and resolution of functional sequence predictions.

Main Methods:

• Developed RELICS, a Bayesian hierarchical model tailored for CRISPR screen data analysis.
• Incorporated methods to handle challenges like unknown functional locations and overdispersion in single-guide RNA counts.
• Designed RELICS to integrate information across multiple experimental pools.

Main Results:

• RELICS demonstrated superior performance over existing methods on simulated datasets, achieving higher precision and recall.
• The model provides finer-resolution predictions of functional sequences.
• Application to published CRISPR interference and activation screens identified novel regulatory sequences missed by other methods.

Conclusions:

• RELICS is a powerful new analysis method for CRISPR screens.
• It enables the discovery of functional sequences with unprecedented resolution and accuracy.
• RELICS facilitates the identification of novel regulatory elements through advanced computational analysis.