cloudrnaSPAdes: isoform assembly using bulk barcoded RNA sequencing data

Area of Science:

• Transcriptomics
• Bioinformatics
• Computational Biology

Background:

• Long-read RNA sequencing offers full-length isoform analysis, but can be low coverage.
• Barcoded short-read RNA sequencing (e.g., 10x Genomics linked-reads) provides higher coverage for isoform studies.
• Existing pipelines struggle with barcoded short-read data, especially with coverage gaps.

Purpose of the Study:

• To present cloudrnaSPAdes, a novel RNA-sequencing assembler.
• To enable the determination of expressed isoforms from barcoded linked-read data in a reference-free manner.
• To overcome limitations of current annotation pipelines for isoform discovery using barcoded short-read RNA-seq.

Main Methods:

• Development of cloudrnaSPAdes, a feature release of the SPAdes assembler.
• Application of the assembler to barcoded RNA-seq linked-read data.
• Reference-free assembly of full-length isoforms.

Main Results:

• cloudrnaSPAdes accurately assembles full-length isoforms from barcoded RNA-seq data.
• The tool demonstrates effectiveness on both simulated and real human datasets.
• Successful isoform determination even for genes with high transcript diversity.

Conclusions:

• cloudrnaSPAdes provides a valuable solution for isoform discovery using high-coverage barcoded RNA-seq data.
• The tool addresses the challenge of analyzing fragmented molecular information from linked-read sequencing.
• This advancement facilitates more comprehensive transcriptomic analysis and isoform annotation.