ARKS

ARKS performs alignment-free scaffolding of genome assemblies using linked read data from 10X Genomics GemCode/Chromium to increase assembly contiguity.

Key Features:

• Alignment-Free Scaffolding: Employs a kmer mapping approach instead of read alignments (e.g., bwa mem) to infer links between reads and assembly contigs, avoiding assembly indexing.
• Kmer Mapping Strategy: Maps barcoded reads via kmers to contig ends to infer connections and construct scaffolds.
• Integration with Hi-C Data: Can be used in conjunction with Hi-C data to further refine scaffolds and improve contiguity metrics such as NG50.
• Improved Performance Metrics: Increased NG50 of a draft human genome assembly from 4.6 Mbp to 23.1 Mbp and improved a Supernova assembly from 14.74 Mbp to 25.94 Mbp, with reported up to ninefold faster runtime than competing linked read scaffolders.
• Efficient Scaffolding: Reduces the number of scaffolds required to cover most chromosomes in tested datasets, with the exception of chromosome 1.
• Novel Distance Estimation: Leverages barcoding information and computes Jaccard indices to model known intra-contig distances and estimate gap sizes between contigs.

Scientific Applications:

• Human Genome Research: Refines human draft genomes toward chromosome-scale assemblies using 10X linked reads.
• Comparative Genomics: Produces higher-contiguity assemblies suitable for cross-species or strain comparisons.
• Functional Genomics: Generates contiguous assemblies that facilitate identification of genes and regulatory elements.

Methodology:

Uses linked read data from 10X Genomics GemCode/Chromium with a kmer-based mapping of barcoded reads to contig ends, computes Jaccard indices between barcode sets, models the relationship between known intra-contig distances to estimate gap sizes, and orders and orients contigs into scaffolds.