Isoform-level gene expression patterns in single-cell RNA-sequencing data.

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for understanding cellular heterogeneity.
• Characterizing isoform-level expression patterns in single cells remains a challenge.
• Existing methods often overlook the complexity of isoform expression variations.

Purpose of the Study:

• To develop and validate a novel computational method, ISOform-Patterns (ISOP), for analyzing single-cell isoform expression.
• To characterize isoform-level expression patterns and heterogeneity in single-cell data.
• To identify genes with differential isoform usage that may be missed by standard differential expression analysis.

Main Methods:

• Developed ISOP, a mixture modeling-based method for isoform-pair analysis in scRNA-seq data.
• Defined six principal patterns of isoform expression relationships.
• Applied ISOP to breast cancer cell line data and validated on three independent datasets.
• Investigated the impact of drop-out events and isoform expression levels using simulated data.

Main Results:

• ISOP successfully assigned expression patterns to 16,562 isoform-pairs across 4929 genes.
• 26% of discovered patterns were statistically significant.
• Differential-pattern analysis identified 32% more genes than traditional differential-expression analysis.
• The method's performance was evaluated concerning drop-out events and expression levels.

Conclusions:

• ISOP offers a novel approach for characterizing isoform-level preference, commitment, and heterogeneity in scRNA-seq data.
• The method enhances the discovery of biologically relevant isoform expression patterns.
• ISOP provides a valuable tool for deeper insights into transcriptional regulation at the single-cell level.