fortuna

fortuna identifies and quantifies alternative splicing events in RNA-seq data using an alignment-free pseudoalignment strategy to detect and annotate known and novel splice junctions.

Key Features:

• Alignment-Free Approach: Employs an alignment-free strategy that avoids full read-to-genome alignment and is reported to be approximately 7× faster than conventional align-and-count approaches.
• Pseudoalignment with kallisto: Uses kallisto pseudoalignment to map reads to generated transcript fragments and to obtain equivalence-class counts.
• Transcript Fragment Generation: Generates transcript fragments by combining annotated splice sites to represent potential novel exons, introns, and junctions.
• Efficient Computation: Demonstrated capability to process ~300 million reads in about 15 minutes using four threads, indicating scalability for large datasets.
• Novel Splicing Event Detection: Detects novel exonic segments and introns, enabling discovery of disease- and tissue-specific splicing events such as those observed in autism spectrum disorder.
• Accurate Mapping Across Novel Junctions: Shows improved accuracy in mapping reads containing mismatches across novel junctions compared to existing methods.
• Tissue and Model-Organism Applicability: Applied to identify tissue-specific splicing events in model organisms, including Drosophila.

Scientific Applications:

• Alternative Splicing Analysis: Quantifies the most elementary splicing units for AS analysis across cell types and tissues.
• Disease Research: Enables detection and characterization of aberrant splicing events implicated in diseases, for example autism spectrum disorder.
• Model Organism Studies: Facilitates discovery of tissue-specific and novel splicing events in model organisms such as Drosophila.

Methodology:

Generates transcript fragments from combinations of annotated splice sites and uses kallisto pseudoalignment to derive equivalence-class counts of basic splicing units that can be aggregated as required.