0
0The paper introduces ROAST, a novel computational tool designed to optimize supertranscriptome assemblies without the need for a reference genome. Creating these assemblies for non-model organisms is often difficult due to errors like fragmented contigs, false chimeras, and local mis-assemblies. Unlike existing methods that rely on external data from related species, ROAST utilizes an iterative algorithm to identify and fix errors solely through RNA-seq alignment signatures. These signatures include soft-clipped bases, inconsistent read coverage, and discordant read pairs from Illumina sequencing. Evaluations using both simulated and real datasets across multiple species demonstrate that this software significantly enhances assembly accuracy and completeness. Ultimately, the tool provides a reference-free solution for refining complex transcriptomic data to support more reliable downstream biological analyses.
References:
Shabbir M, Mithani A. Roast: a tool for reference-free optimization of supertranscriptome assemblies. BMC Bioinformatics. 2024 Jan 2;25(1):2. doi: 10.1186/s12859-023-05614-4. PMID: 38166712; PMCID: PMC10763045.
