flipflop

flipflop identifies and quantifies transcript isoforms from RNA-Seq read alignments (SAM) using network flow optimization to enable accurate sparse estimation of isoform abundances.

Key Features:

• Isoform Identification and Quantification: Determines which isoforms are expressed in a sample and quantifies their abundances from RNA-Seq alignments.
• Input Format: Accepts alignment files in SAM format.
• Simultaneous Multi-Sample Analysis: Processes multiple samples concurrently to support comparative analyses across conditions or tissues.
• Network Flow Optimization: Employs network flow optimization to perform sparse estimation across the full set of candidate transcripts without requiring preselection.
• Alternative to Regularized Regression: Uses network flow optimization instead of [Formula: see text]-regularized regression (e.g., Lasso) to address combinatorial explosion of candidate transcripts in genes with many exons.
• Scalability and Efficiency: Handles genes with numerous exons with reduced computational cost by avoiding preselection of isoforms.
• Performance Validation: Evaluated on synthetic and real RNA-Seq datasets and reported to outperform alternative methods in accuracy and speed.

Scientific Applications:

• Gene Expression Studies: Investigating differential splicing and isoform-level expression across conditions or developmental stages.
• Disease Research: Identifying isoform-specific expression patterns associated with diseases for biomarker or therapeutic target discovery.
• Comparative Genomics: Analyzing isoform diversity and evolution across species.

Methodology:

Performs sparse estimation of isoform abundances using network flow optimization across the full set of candidate transcripts as an alternative to [Formula: see text]-penalized regression (e.g., Lasso).