DTA-SiST: de novo transcriptome assembly by using simplified suffix trees.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Alternative splicing generates diverse proteins from single genes.
• High-throughput sequencing enables transcript reconstruction.
• Short sequencing reads pose algorithmic challenges for de novo transcriptome assembly.

Purpose of the Study:

• To develop a novel framework for de novo transcriptome assembly.
• To address the algorithmic challenges posed by short sequencing reads.
• To improve the accuracy and efficiency of transcript reconstruction.

Main Methods:

• Developed DTA-SiST, a de novo transcriptome assembly framework utilizing suffix trees.
• Implemented a read-based contig extension strategy using longest overlaps.
• Constructed gene-specific splicing graphs.
• Employed depth-first enumeration and hybrid strategies for transcript path extraction.

Main Results:

• DTA-SiST efficiently extends contigs using a suffix tree structure.
• The depth-first strategy excels in recall and precision on smaller datasets.
• The hybrid strategy achieves superior precision on larger datasets.
• Comparative analyses on simulated and real data demonstrate DTA-SiST's competitiveness.

Conclusions:

• DTA-SiST offers a competitive alternative to existing de novo assemblers.
• The read-based contig extension and transcript extraction rules enhance assembly precision.
• The framework effectively reconstructs transcripts from short-read sequencing data.