MINTIE

MINTIE (Method for Identifying Novel Transcripts and Isoforms using Equivalence classes) identifies novel and rare cryptic transcript variants from RNA-seq data to detect disease-specific transcriptional changes, including up-regulated novel isoforms and variants beyond fusion genes.

Key Features:

• Reference-free approach: Integrates de novo transcript assembly with differential expression analysis without relying on pre-existing genomic references.
• De novo transcript assembly: Assembles transcripts directly from RNA-seq data to enable discovery of novel isoforms and transcript structures.
• Differential expression analysis: Detects up-regulated novel variants by comparing case samples to controls.
• Equivalence classes: Operates using equivalence classes as indicated by the method name to support transcript-level analysis.
• Detection beyond fusion genes: Identifies novel and rare cryptic variants that are not limited to fusion events.
• Benchmark performance: Was compared against eight other approaches and detected over 85% of transcript variants in that comparison.
• Input data: Designed to analyze RNA-seq datasets.

Scientific Applications:

• Cancer genomics: Identification of novel transcript variants and isoforms in cancer RNA-seq to reveal tumor-specific transcriptional changes.
• Rare disease research: Detection of rare cryptic RNA variants that may contribute to genetic disease phenotypes.
• Biomarker and target discovery: Discovery of candidate diagnostic markers or therapeutic targets through identification of disease-associated novel transcripts.
• Transcriptome variation studies: Exploration of transcriptomic variation and up-regulated novel isoforms in diseased tissues.

Methodology:

MINTIE applies a reference-free pipeline that combines de novo transcript assembly with differential expression analysis and uses equivalence classes; its performance was benchmarked against eight other methods with reported detection of over 85% of transcript variants.