ganon2: up-to-date and scalable metagenomics analysis

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Public genomic sequence repositories are growing rapidly, outpacing available computational resources.
• Current metagenomic analysis methods struggle to efficiently process massive, fast-generating datasets.

Purpose of the Study:

• To introduce ganon2, a novel sequence classification method for metagenomics analysis.
• To enhance performance and usability in taxonomic binning and profiling.
• To enable better utilization of large and up-to-date reference datasets.

Main Methods:

• ganon2 utilizes a novel indexing approach for large datasets with a small memory footprint.
• The method performs taxonomic binning and profiling for metagenomics.
• Indices were built using the NCBI RefSeq database and its subsets.

Main Results:

• ganon2 indices are 50% smaller on average compared to state-of-the-art methods.
• Achieved up to 0.15 higher median F1-score in taxonomic binning across simulated samples.
• Showed up to 0.35 higher median F1-score in profiling with balanced L1-norm error.

Conclusions:

• ganon2 provides a generational leap in performance and usability for metagenomic analysis.
• The tool enables the use of larger, more diverse, and current reference sets, improving result resolution.
• ganon2 is a fast, open-source tool for efficient taxonomic classification in microbiome studies.