Dissect: detection and characterization of novel structural alterations in transcribed sequences

Area of Science:

• Bioinformatics
• Genomics
• Transcriptomics

Background:

• High-throughput sequencing enables genomic structural variant identification.
• Transcriptome sequencing (RNA-Seq) is increasingly used to detect transcriptomic alterations.
• Existing methods for transcriptomic structural variant discovery are limited.

Purpose of the Study:

• Introduce novel algorithmic formulations for identifying transcriptomic structural variants.
• Present Dissect (DIScovery of Structural Alteration Event Containing Transcripts), a new tool for transcriptome-to-genome alignment.
• Enable the characterization of transcriptomic events like duplications, inversions, rearrangements, and fusions.

Main Methods:

• Developed two algorithmic formulations for transcriptomic structural variant identification.
• Nucleotide-level alignment model.
• Fragment chaining model for potentially faster analysis.
• Implemented Dissect tool for whole transcriptome structural variation discovery.

Main Results:

• Dissect demonstrated high sensitivity and specificity on simulated data.
• The tool successfully identified novel structural alterations in human prostate cancer cell line C4-2 RNA-Seq data.
• Dissect is effective for long reads or accurately assembled contigs.

Conclusions:

• Dissect is a powerful tool for discovering transcriptomic structural variants.
• The algorithms and tool advance the field of cancer transcriptome analysis.
• Dissect is publicly available for research use.