SCRABBLE: single-cell RNA-seq imputation constrained by bulk RNA-seq data.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) data frequently exhibit a high proportion of zero counts, known as dropout events.
• These dropout events pose a significant challenge for downstream data analysis and interpretation.
• Accurate gene expression quantification is crucial for understanding cellular heterogeneity and function.

Purpose of the Study:

• To introduce and validate SCRABBLE, a novel algorithm designed to address zero-inflation in scRNA-seq data.
• To improve the accuracy of gene expression imputation by leveraging bulk RNA-seq data as a constraint.
• To reduce bias and enhance the reliability of scRNA-seq data analyses.

Main Methods:

• SCRABBLE utilizes bulk RNA-seq data to inform and constrain the imputation process for scRNA-seq data.
• The algorithm is designed to minimize bias towards highly expressed genes during imputation.
• Performance was evaluated using both simulated datasets and diverse experimental scRNA-seq data.

Main Results:

• SCRABBLE demonstrated superior performance compared to existing methods in recovering dropout events.
• The algorithm accurately captured the true distribution of gene expression across individual cells.
• SCRABBLE effectively preserved both gene-gene and cell-cell relationships within the dataset.

Conclusions:

• SCRABBLE offers a robust solution for handling zero counts in scRNA-seq data.
• The algorithm enhances the accuracy of gene expression imputation and preserves critical biological signals.
• SCRABBLE is a valuable tool for advancing the analysis of single-cell transcriptomics.