scDecouple: decoupling cellular response from infected proportion bias in scCRISPR-seq.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Single-cell clustered regularly interspaced short palindromic repeats-sequencing (scCRISPR-seq) is a powerful tool for high-throughput screening.
• Analyzing scCRISPR-seq data is challenging due to the confounding effects of true cellular responses and infected proportion bias across cell clusters.

Purpose of the Study:

• To develop a computational method, scDecouple, to accurately distinguish true cellular responses from infected proportion bias in scCRISPR-seq data.
• To improve the identification of biologically relevant genes affected by perturbations.

Main Methods:

• scDecouple models gene expression distributions in perturbed cells.
• It employs iterative maximum likelihood estimation to decouple cell cluster proportions and cellular responses for each guide RNA (gRNA).
• The method was validated using simulation experiments and real scCRISPR-seq datasets.

Main Results:

• scDecouple effectively separates true biological signals from technical noise in scCRISPR-seq data.
• Application to real data demonstrated enhanced identification of perturbation-related genes.
• The method proved robust in simulation studies.

Conclusions:

• scDecouple offers a significant advancement for scCRISPR-seq data analysis.
• It is particularly beneficial for heterogeneous samples and complex gRNA libraries.
• This tool facilitates more accurate biological insights from CRISPR screening experiments.