SCReadCounts: estimation of cell-level SNVs expression from scRNA-seq data.

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) enables detailed analysis of cellular heterogeneity and gene expression.
• Single nucleotide variants (SNVs) are crucial for distinguishing cell types and understanding biological processes like transcriptional burst kinetics and allelic expression.
• Existing methods require robust tools for accurate SNV analysis at the single-cell level.

Purpose of the Study:

• To develop SCReadCounts, a computational tool for tabulating SNV read counts from scRNA-seq data.
• To provide cell-SNV matrices for variant allele fraction (VAF_RNA) analysis.
• To enable downstream applications such as somatic mutation detection and allele-specific expression analysis.

Main Methods:

• Developed SCReadCounts for cell-level tabulation of reference and variant allele read counts from barcoded scRNA-seq alignments.
• Generated cell-SNV matrices of absolute read counts and VAF_RNA.
• Applied SCReadCounts to 59,884 cells from seven neuroblastoma samples for mutation expression, biallelic SNV expression, and SNV discovery.
• Benchmarked SCReadCounts against GATK and Samtools.

Main Results:

• SCReadCounts successfully generated cell-SNV matrices for various downstream analyses.
• Demonstrated applications in estimating cell-level expression of somatic mutations and RNA-editing sites.
• Enabled cell-level allele expression analysis of biallelic SNVs.
• Identified known and novel somatic mutations in KRAS using discovery mode.

Conclusions:

• SCReadCounts offers a fast and efficient solution for estimating cell-level SNV expression from scRNA-seq data.
• The tool can distinguish cells with monoallelic reference expression from those with no gene expression.
• SCReadCounts is applicable for assessing SNVs present in a small proportion of cells, including somatic mutations in cancer.