SCISSOR

SCISSOR detects structural alterations in RNA transcripts from RNA-seq by analyzing base-level coverage profiles to enable unsupervised discovery of aberrant transcript shapes.

Key Features:

• Base-level coverage analysis: Analyzes RNA-seq coverage profiles at base-level resolution to detect shape changes in expression profiles.
• Short read transformation and normalization: Transforms and normalizes short read sequencing data into an intact and uncompressed view for analysis.
• Annotation-independent detection: Operates independently of specific transcript information or annotations.
• High-dimensional representation: Converts raw sequencing data into high-dimensional objects suitable for statistical analysis.
• Unsupervised statistical analysis: Performs unsupervised detection of anomalous shapes without assuming underlying mutational mechanisms.
• Variant types detected: Identifies abnormal splicing events, intra- and intergenic deletions, small insertions and deletions (indels), and alternative transcription start and termination sites.
• Recapitulation of known variants: Capable of recapturing known variants such as splice site mutations in tumor suppressor genes.
• Discovery of complex events: Uncovers recurrent alternate transcription start sites and complex deletions within 3' untranslated regions (UTRs).

Scientific Applications:

• Aberrant splicing detection: Identifies abnormal splicing events from anomalous coverage shapes in RNA-seq data.
• Deletion and indel discovery: Detects intra- and intergenic deletions and small insertions and deletions (indels) affecting transcript structure.
• Transcription start/termination discovery: Discovers alternative transcription start sites and alternative termination sites from coverage profile changes.
• Variant recovery in cancer genes: Recovers known splice site mutations in tumor suppressor genes from RNA-seq coverage patterns.
• 3' UTR structural variation discovery: Identifies complex deletions and recurrent alternate events within 3' untranslated regions (UTRs).
• Transcriptome-wide structural screening: Enables unbiased, transcript-annotation-independent screening for structural alterations across RNA-seq datasets.

Methodology:

Analyzes RNA-seq base-level coverage profiles, transforms and normalizes short read sequencing data into an intact uncompressed representation, converts these into high-dimensional objects, and applies unsupervised statistical analysis to identify anomalous expression-profile shapes without assuming specific mutational mechanisms.