The eSNV-detect: a computational system to identify expressed single nucleotide variants from transcriptome...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing (NGS) technologies facilitate genomic alteration identification from short reads.
• Existing pipelines primarily focus on genomic DNA variants, lacking comprehensive solutions for expressed single nucleotide variants (eSNVs) from RNA sequencing (RNA-Seq) data.

Purpose of the Study:

• To develop a novel computational system, eSNV-Detect, for identifying, annotating, and prioritizing eSNVs from non-directional paired-end RNA-Seq data.
• To evaluate the performance of eSNV-Detect across multiple platforms and datasets.

Main Methods:

• eSNV-Detect utilizes data from multiple aligners to call and rank variants from RNA-Seq, even at low read depths.
• The system was validated using RNA-Seq data from a lymphoblastoid cell line, breast tumors (TCGA), and single-cell mRNA-Seq.
• Sanger sequencing was employed for further validation of candidate eSNVs.

Main Results:

• eSNV-Detect achieved 99.7% precision and 91.0% sensitivity for expressed SNPs in lymphoblastoid cell line data.
• Comparisons with whole exome coding data from breast tumors showed 90.6-96.8% precision and 91.6-95.7% sensitivity.
• Analysis of single-cell data revealed variant heterogeneity, and Sanger sequencing validated 29 out of 31 candidate eSNVs.

Conclusions:

• eSNV-Detect is a comprehensive and accurate pipeline for identifying and prioritizing expressed single nucleotide variants from RNA-Seq data.
• The system demonstrates high performance across diverse biological samples, including cancer.
• eSNV-Detect provides valuable insights into variant heterogeneity at the single-cell level.