Ganon

Ganon performs k-mer-based read classification using Interleaved Bloom Filters and taxonomic clustering to classify metagenomic sequencing reads against large, frequently updated reference databases.

Key Features:

• Efficient Indexing: Indexing uses Interleaved Bloom Filters combined with taxonomic clustering and a k-mer counting/filtering scheme to enable rapid indexing of large reference sets.
• Reference Scope and Indexing Performance: Demonstrated indexing of complete RefSeq genomes (bacteria, archaea, fungi, viruses) in under 55 minutes on standard servers.
• Incremental Index Updates: Indices can be maintained and updated more quickly than full re-creation, enabling frequent incorporation of new reference sequences.
• Classification Algorithms: Performs k-mer-based read classification and supports taxonomy-level and assembly-level classification.
• Multi-index and Hierarchical Classification: Supports handling multiple indices simultaneously and provides hierarchical classification options.
• Benchmark Performance: On a high-complexity CAMI challenge dataset against complete RefSeq genomes, achieves higher precision and equal or better sensitivity with a reported ~65% improvement in F1-score at the genus level.

Scientific Applications:

• Metagenomic taxonomic profiling: Classifies sequencing reads to taxonomic or assembly levels to support profiling of microbial communities.
• Microbial diversity studies: Enables identification of a greater number of species and more accurate read classification in diversity analyses.
• Pathogen detection: Supports detection and classification of pathogenic taxa from metagenomic sequence data.
• Environmental monitoring: Facilitates analysis of environmental metagenomes requiring up-to-date reference genomes.

Methodology:

Indexing with Interleaved Bloom Filters combined with taxonomic clustering and a k-mer counting/filtering scheme; k-mer-based read classification; support for multiple indices, hierarchical classification, and incremental index updates.