TranD

TranD implements a mathematical framework to calculate complexity and compare transcript models by quantifying alternative splicing at the nucleotide level.

Key Features:

• Nucleotide-level distance metrics: Calculates distance between transcript models with nucleotide-level precision.
• Structural feature comparison: Compares intron retention, donor/acceptor site variation, alternative exon cassettes, and alternative 5'/3' UTRs.
• 1GTF mode: Summarizes splicing patterns for a single transcriptome.
• 2GTF mode: Compares splicing patterns between two transcriptomes.

Scientific Applications:

• Annotation validation: Quantitatively compares annotations augmented by empirical RNA-seq data against original transcript models.
• Transcript prediction tool benchmarking: Compares transcript model prediction tools for long-read RNA-seq such as FLAIR versus Isoseq3.
• Cross-annotation comparison: Compares alternate annotations for a species such as RefSeq versus Ensembl.
• Cross-species comparison: Compares transcript models between closely related species such as D. melanogaster and D. simulans.

Methodology:

TranD implements a mathematical framework to calculate complexity and compare transcript models, with all metrics packaged as a PyPi package.