DTA-SiST

DTA-SiST reconstructs de novo transcriptomes from RNA-seq short reads using suffix trees to assemble contigs and identify transcript-representing paths, enabling analysis of gene expression and alternative splicing.

Key Features:

• Suffix Tree-Based Contig Extension: Employs a suffix tree to extend contigs by identifying reads with the longest terminal overlaps in linear time relative to read length.
• Splicing Graph Construction: Constructs splicing graphs for each gene locus from extended contigs to represent exon connectivity and alternative splicing.
• Depth-First Enumeration Strategy: Systematically explores all paths in the splicing graph to enumerate candidate transcripts, prioritizing recall and precision for smaller datasets.
• Hybrid Strategy: Combines length and coverage metrics to select transcript candidates and optimize precision for larger datasets.

Scientific Applications:

• De novo transcriptome reconstruction: Reconstructs transcript sequences from RNA-seq data without a reference genome.
• Alternative splicing analysis: Identifies transcript isoforms by analyzing splicing graph paths derived from contigs.
• Gene expression profiling: Provides assembled transcripts for downstream quantification of gene and isoform expression.
• Functional annotation: Supplies transcript sequences for downstream functional annotation of predicted isoforms.

Methodology:

Contig extension using suffix trees to merge overlapping reads; splicing graph construction for each gene locus from extended contigs; transcript extraction via depth-first enumeration and a hybrid length-and-coverage strategy.