Fusion transcripts and their genomic breakpoints in polyadenylated and ribosomal RNA-minus RNA sequencing data

Area of Science:

• Genomics
• Transcriptomics
• Bioinformatics

Background:

• Fusion genes are typically identified using RNA sequencing (RNA-seq) without pinpointing causal genomic breakpoints.
• Non-poly(A)-enriched RNA-seq data contains intronic reads that can span genomic breakpoints.

Purpose of the Study:

• To develop and validate an algorithm for identifying expressed genomic breakpoints from RNA-seq data.
• To investigate the utility of non-poly(A)-enriched RNA-seq for detecting fusion transcripts and their implications.

Main Methods:

• Development of the Dr. Disco algorithm, which searches the entire reference genome, including introns and intergenic regions, for fusion transcripts.
• Analysis of 1,275 RNA-seq samples to compare breakpoint detection in poly(A)-enriched versus non-poly(A)-enriched RNA-seq data.
• Comparison with whole-genome sequencing data to assess the transcription of genomic breakpoints.

Main Results:

• Dr. Disco successfully identified expressed genomic breakpoints from RNA-seq data.
• Most genomic breakpoints are minimally transcribed, but breakpoints in TMPRSS2-ERG-positive tumors were consistently detected at the RNA level.
• Identified novel fusion transcripts involving intergenic regions and cryptic exons, potentially encoding neo-antigens, and linked rearrangement hotspots to increased gene expression in breast and brain cancers.

Conclusions:

• Sophisticated analysis of non-poly(A)-enriched RNA-seq data can reliably identify expressed genomic breakpoints and their transcriptional consequences.
• The Dr. Disco algorithm expands the scope of fusion transcript detection beyond classical gene-to-gene fusions.