Cell-level somatic mutation detection from single-cell RNA sequencing.

Area of Science:

• Genomics
• Computational Biology
• Cancer Research

Background:

• Single-cell sequencing technologies like scRNA-seq and scDNA-seq offer valuable cell-level genomic profiling.
• Detecting mutations from scRNA-seq data is challenging due to limited input material and amplification biases, unlike scDNA-seq.
• Existing mutation detection methods for bulk or scDNA-seq data perform poorly on scRNA-seq data, yielding many false positives.

Purpose of the Study:

• To develop a robust statistical method for accurate cell-level mutation detection using scRNA-seq data.
• To address the limitations of existing mutation detection tools when applied to scRNA-seq.
• To enable the investigation of cell-to-cell heterogeneity through precise mutation identification.

Main Methods:

• Development of SCmut, a novel statistical method for identifying specific cells harboring mutations from bulk-cell data.
• Implementation of a 2D local false discovery rate method within SCmut for robust false positive control.
• Application and validation of SCmut on multiple scRNA-seq datasets, including breast cancer and glioblastoma.

Main Results:

• SCmut successfully identifies highly confident cell-level mutations from scRNA-seq data, showing recurrence across many cells and consistency between samples.
• In breast cancer scRNA-seq datasets, SCmut detected recurrent cell-level mutations.
• In a glioblastoma scRNA-seq dataset, SCmut discovered a recurrent mutation in the PDGFRA gene, correlated with a known deletion.

Conclusions:

• SCmut provides a novel and effective approach for discovering cell-level mutation information from scRNA-seq data.
• This method facilitates the investigation of cell-to-cell heterogeneity by accurately identifying mutations at the single-cell level.
• The developed bioinformatics pipeline and source code are publicly available for broader research application.