McCortex

McCortex augments de Bruijn graphs with long-range connectivity annotations to preserve sequence context for genome assembly, read error correction, and variant calling.

Key Features:

• Linked De Bruijn Graph Structure: Implements the Linked de Bruijn Graph (LdBG) data structure that augments standard de Bruijn graphs with annotations storing long-range connectivity derived from sequencing reads.
• Lossless Data Representation: Can losslessly store and recover sequences from error-free data using the LdBG representation.
• Performance Superiority: Demonstrated improved assembly performance in simulations compared to traditional de Bruijn graphs and the String Graph Assembler (SGA) by leveraging long-range connectivity.
• Variant Calling and Genomic Characterization: Applied to short-read data from Klebsiella pneumoniae to identify large variants up to 12 kilobases (kbp), validated with PacBio sequencing, and to characterize genomic contexts around drug-resistance genes.

Scientific Applications:

• Genome Assembly: Improving reconstruction of genomes from short-read sequencing by preserving long-range read connectivity.
• Read Error Correction: Enhancing read error correction by using long-range connectivity annotations.
• Variant Calling: Detecting genetic variants including large structural variations up to 12 kilobases (kbp).
• Microbial Genomics: Characterizing genomic features related to antibiotic resistance, demonstrated on Klebsiella pneumoniae.
• High-depth and Multi-sample Projects: Supporting analyses of large genomes with high sequencing depth and multiple samples.

Methodology:

Algorithms construct and manipulate the Linked de Bruijn Graph (LdBG) to integrate long-range connectivity annotations into the de Bruijn graph framework.