Unsupervised correction of gene-independent cell responses to CRISPR-Cas9 targeting

Area of Science:

• Genomics
• Cancer Biology
• Bioinformatics

Background:

• CRISPR-Cas9 technology enables large-scale gene essentiality screening in cancer.
• High false-positive rates in CRISPR-Cas9 screens, particularly in amplified genomic regions, confound essential gene identification.
• Gene-independent responses to CRISPR-Cas9 targeting present a challenge in data interpretation.

Purpose of the Study:

• To develop a computational tool, CRISPRcleanR, for identifying and correcting gene-independent responses in CRISPR-Cas9 screening data.
• To improve the accuracy of essential gene detection in cancer by mitigating biases in CRISPR-Cas9 screens.

Main Methods:

• CRISPRcleanR employs an unsupervised approach using genome-wide segmentation of single-guide RNA fold change values.
• The method does not assume any specific copy number status of the targeted genes.
• It analyzes genome-wide essentiality profiles from cancer cell lines.

Main Results:

• CRISPRcleanR effectively reduces false positives in essential gene calling.
• The tool corrects biases in CRISPR-Cas9 screening data, both within and outside amplified genomic regions.
• True positive rates are maintained, and established cancer dependencies are detectable post-correction.

Conclusions:

• CRISPRcleanR is a versatile, open-source tool for analyzing CRISPR-Cas9 knockout screens.
• It enhances the identification of essential genes in cancer research.
• The tool supports downstream analysis and is compatible with various sgRNA libraries.