AIDE

AIDE identifies and quantifies full-length mRNA isoforms from second-generation short-read RNA-seq data for genome-wide isoform discovery and abundance estimation.

Key Features:

• Annotation-Assisted Approach: Integrates existing annotated isoform data and performs stepwise discovery that prioritizes and selectively borrows information from known annotations using a likelihood ratio test.
• Statistical Rigor: Implements a testing-based model selection principle to directly control false discoveries and retains novel isoforms only when their inclusion significantly improves the explanation of observed RNA-seq reads.
• High Precision and Low Error Rates: Comparative evaluations on simulated and real RNA-seq datasets, validated by PCR-Sanger sequencing, report higher precision in isoform discovery and lower error rates in abundance estimation than Cufflinks, SLIDE, and StringTie.
• Robustness for Transcriptome Analysis: Enables confident discovery of novel transcripts to support analyses of transcriptional and posttranscriptional regulatory mechanisms.

Scientific Applications:

• Transcriptional Regulation: Identification of novel and known isoforms to study how transcription factors and regulatory elements shape gene expression diversity.
• Posttranscriptional Modifications: Characterization of alternative splicing and other RNA processing events that generate transcript and protein diversity.
• Disease Mechanisms: Detection and quantification of aberrant isoforms relevant to genetic diseases and conditions involving altered splicing or transcriptional regulation.

Methodology:

Integrates annotated isoform data with observed RNA-seq reads and applies a likelihood ratio test within a stepwise, annotation-prioritized model selection framework that directly controls false discoveries.