Hybracter

Hybracter assembles accurate, near-complete bacterial genomes from hybrid (short- and long-read) and long-read sequencing data to enable reconstruction of plasmids and analysis of genomic variation, including plasmid-borne antimicrobial resistance.

Key Features:

• Hybrid assembly: Integrates both short- and long-read sequencing data to reconstruct comprehensive bacterial genomic sequences.
• Long-read only assembly: Can operate using long-reads alone as an alternative assembly mode.
• Long-read-first strategy: Implements a long-read-first assembly approach intended to enhance recovery of plasmids, including small plasmids.
• Small plasmid recovery: Improves detection and accurate assembly of small plasmids that are often missed or misassembled by other long-read algorithms.
• Accuracy and completeness: Produces near-perfect complete bacterial genome assemblies with accuracy comparable to hybrid approaches.
• Performance benchmarking: Demonstrates superior accuracy and speed compared to existing automated hybrid tools such as Unicycler and versus other long-read only methods.
• Scalability and automation: Provides scalable, automated assembly for analysis of multiple bacterial samples.
• Validated on diverse datasets: Benchmarked on a diverse panel of samples with varying long-read accuracy against manually curated ground truth reference genomes.

Scientific Applications:

• Complete genome reconstruction: Reconstruction of complete bacterial chromosomes and plasmids for genomic analyses.
• Plasmid identification and analysis: Recovery and characterization of plasmids to study their structure and gene content.
• Antimicrobial resistance studies: Identification and analysis of plasmid-borne antimicrobial resistance genes to investigate dissemination mechanisms.
• Bacterial evolution and variation: Investigation of genomic variation beyond single nucleotide variants to study bacterial evolution.
• Benchmarking and method comparison: Comparative evaluation of assembly approaches using curated reference genomes.

Methodology:

Performs hybrid assembly by integrating short- and long-read data, offers a long-read-first assembly strategy and a long-read-only mode, and has been compared against Unicycler and other long-read only methods and validated on a diverse panel against manually curated ground truth reference genomes.