Fast and sensitive taxonomic assignment to metagenomic contigs

Area of Science:

• Bioinformatics
• Computational Biology
• Metagenomics

Background:

• Accurate taxonomic classification of metagenomic contigs is crucial for understanding microbial community structure and function.
• Existing tools may face limitations in speed, scalability, or comprehensiveness for diverse genomic data.

Purpose of the Study:

• Introduce MMseqs2 taxonomy, a new computational tool for efficient and robust taxonomic assignment of metagenomic contigs.
• Enhance the analysis of all domains of life through advanced fragment extraction and weighted voting strategies.
• Provide integrated modules for reference database management and visualization of taxonomic data.

Main Methods:

• Extraction of all protein fragments from each metagenomic contig.
• Filtering and annotation of fragments using robust labeling.
• Weighted voting for determining the final taxonomic identity of each contig.
• Development of modules for reference database creation, manipulation, reporting, and visualization.

Main Results:

• MMseqs2 taxonomy achieves 2-18x speed improvement over state-of-the-art tools.
• The fragment extraction method is applicable across all domains of life.
• New modules facilitate comprehensive management and interpretation of taxonomic assignments.

Conclusions:

• MMseqs2 taxonomy provides a significantly faster and more versatile solution for metagenomic taxonomic annotation.
• The tool's design supports broad applicability and enhances the analysis of complex microbial datasets.
• Integrated database and reporting features streamline the entire taxonomic profiling workflow.