RResolver

RResolver improves de novo genome assembly by optimizing de Bruijn graph resolution of repeat regions in short-read sequencing data.

Key Features:

• De Bruijn Graph Optimization: Operates on a de Bruijn graph produced by ABySS, using k-mers and adjusting the k value toward the read length to improve resolution of repeat regions.
• Bloom Filter Integration: Employs a Bloom filter of sequencing reads to rapidly assess path support at branching points and remove paths with insufficient support.
• Scalability and Performance: Demonstrated assembly of a human genome with ABySS in approximately 26 minutes using 48 threads and 60 GiB of memory, enabling processing of large genomes.
• Improved Assembly Outcomes: Increases scaffold contiguity (NGA50) by up to 15% and reduces misassemblies by up to 12%.
• Integration with ABySS: Functions within the ABySS assembly pipeline to enhance graph traversal and provide a simplified genomic representation for downstream algorithms.
• Targeting High-Variation Genomes: Addresses assembly challenges in highly variable genomes, such as those encountered in cancer studies, by optimizing repeat resolution in short-read data.

Scientific Applications:

• Cancer genomics: Improves de novo assemblies of highly variable cancer genomes to support variant and structural analyses.
• Large-genome de novo assembly: Enhances contiguity and accuracy of de novo assemblies for large genomes generated from short-read sequencing and ABySS.
• Downstream genomic analyses: Produces more contiguous and accurate assemblies that facilitate downstream analyses requiring high scaffold continuity.

Methodology:

Operates on an ABySS-derived de Bruijn graph built from k-mers of short reads, adjusts k toward read length to resolve repeats, and uses a Bloom filter of sequencing reads to evaluate path support at branching points and remove unsupported paths.