TALC

TALC performs reference-free, transcription-aware correction of long-read RNA sequencing data to improve transcriptome reconstruction and isoform representation.

Key Features:

• Transcription-Aware Correction: Corrects long reads using information specific to transcriptome data to account for splicing and isoform variability.
• Reference-Free Approach: Operates without requiring a pre-existing reference genome or transcriptome.
• Weighted De-Bruijn Graph Model: Uses a weighted De-Bruijn graph to represent RNA expression changes and isoform representation during correction.
• Alternative to Hybrid Correction: Does not rely on short genomic reads for hybrid correction, focusing instead on transcriptome-derived signals.
• Implementation: Implemented in C++.

Scientific Applications:

• Transcriptome Assembly: Improves accuracy of assembled transcriptomes from long-read RNA-seq data.
• Isoform Detection: Enhances detection and resolution of alternative splicing events and transcript isoforms.
• Expression Quantification: Increases reliability of gene and isoform expression estimates from long reads.

Methodology:

Reference-free algorithm that models RNA expression changes and isoform representation within a weighted De-Bruijn graph to distinguish sequencing errors from true biological variation.