DETONATE

DETONATE evaluates de novo transcriptome assemblies and provides model-based and ground-truth-based metrics to assess assembly accuracy from RNA-Seq data.

Key Features:

• RSEM-EVAL: A model-based scoring system that evaluates RNA-Seq assemblies without requiring known ground truth by using statistical models to estimate assembly accuracy.
• REF-EVAL: A ground-truth-based scoring component that computes refined scores to compare sets of genomic sequences, including de novo transcriptome assemblies.
• Assembly comparison: Produces quantitative metrics for comparing different transcriptome assemblies to guide selection of the most accurate assembly.
• Benchmarking and validation: Uses REF-EVAL ground-truth scores to validate and contextualize RSEM-EVAL model-based assessments.
• Applicability to species without reference genomes: Designed to evaluate assemblies from species lacking sequenced genomes using RNA-Seq data.

Scientific Applications:

• De novo transcriptome evaluation: Assess the quality and accuracy of de novo transcriptome assemblies generated from RNA-Seq experiments.
• Assembly selection: Compare and rank alternative transcriptome assemblies to identify the most reliable reconstruction.
• Metric validation: Validate model-based evaluation metrics against ground-truth-based scores to ensure consistency in performance assessment.
• Axolotl limb regeneration transcriptome: Applied to assembling and assessing the regenerating axolotl limb transcriptome.

Methodology:

Applies statistical model-based scoring (RSEM-EVAL) to RNA-Seq assemblies without ground truth and computes ground-truth-based scoring metrics (REF-EVAL) to compare genomic sequence sets and validate assembly accuracy.