GoldRush

GoldRush performs de novo assembly of long-read sequencing data using a dynamic data structure to avoid all-vs-all read overlaps and attain linear time complexity.

Key Features:

• Linear Time Complexity: Achieves linear time complexity by avoiding computationally intensive all-vs-all sequence alignments.
• Dynamic Data Structure: Employs a novel dynamic data structure to manage and assemble long reads without performing all-vs-all read-to-read overlaps.
• Reduced Memory Usage and Scalability: Assembles human, rice, and tomato genomes within a day while consuming at most 54.5 GB of RAM.
• Compatibility with Oxford Nanopore Data: Optimized for Oxford Nanopore Technologies long-read sequencing datasets and reported to handle varying base error profiles.
• Assembly Quality Metrics: Produces scaffold NGA50 lengths of 18.3–22.2 Mbp for human, 0.3 Mbp for rice, and 2.6 Mbp for tomato.

Scientific Applications:

• De novo genome assembly: Generation of genome assemblies from Oxford Nanopore Technologies long-read sequencing datasets.
• Assemblies for diverse organisms: Applied to human cell lines, rice, and tomato genomes.
• Large-scale projects with limited resources: Enables scalable de novo assembly workflows for projects constrained by memory and runtime.

Methodology:

Foregoes read-to-read all-vs-all overlap computation and leverages a dynamic data structure to assemble long reads, reducing computational overhead.