ChopStitch: exon annotation and splice graph construction using transcriptome assembly and whole genome sequencing...

Area of Science:

• Bioinformatics
• Genomics
• Transcriptomics

Background:

• De novo sequencing studies on non-model organisms generate massive short sequence data.
• High-quality reference genomes are often unavailable for these species.
• Accurate de novo annotation and alternative splicing detection are crucial for understanding gene function.

Purpose of the Study:

• To present ChopStitch, a novel computational method for de novo exon identification and splice graph construction.
• To enable the analysis of transcriptomes and whole genome shotgun sequencing data for non-model organisms.
• To facilitate de novo annotation and the discovery of transcript isoforms.

Main Methods:

• ChopStitch utilizes de novo assembled RNA-Seq data and whole genome shotgun sequencing (WGSS) data.
• A Bloom filter representing the WGSS k-mer spectrum aids in identifying exon-exon boundaries.
• The algorithm accounts for sequence variations including base substitutions, haplotype variations, and RNA editing.

Main Results:

• The primary output is a FASTA file of putative exons.
• Splice graphs in DOT format represent transcript isoforms by interrogating alternative exon-exon boundaries.
• Enables de novo annotation and isoform detection without high-quality reference genomes.

Conclusions:

• ChopStitch provides a robust solution for analyzing sequencing data from non-model organisms.
• The tool facilitates the identification of exons and splice variants, crucial for genomic annotation.
• It supports research in species lacking extensive genomic resources.