AKE - the Accelerated k-mer Exploration web-tool for rapid taxonomic classification and visualization

Area of Science:

• Computational Biology
• Bioinformatics
• Genomics

Background:

• Advancements in low-cost, high-throughput sequencing enable large-scale metagenomic analyses.
• Handling the vast data volumes and inherent diversity in metagenomic datasets presents a significant computational challenge.

Purpose of the Study:

• To develop a method for rapid taxonomic assignment in metagenomics.
• To introduce the Accelerated k-mer Explorer (AKE) with adaptive, small data models (< 5 MB).

Main Methods:

• Utilized a specialized machine learning architecture designed for hierarchical data.
• Implemented the Accelerated k-mer Explorer (AKE) for taxonomic classification.
• Tested on real-world metagenomic datasets, including Acid Mine Drainage and Cow Rumen samples.

Main Results:

• Achieved classification accuracy down to the order rank.
• Demonstrated that AKE's taxonomic assignments are significantly faster than existing methods.
• Maintained comparable accuracy to other approaches.

Conclusions:

• AKE offers a computationally efficient solution for metagenomic taxonomic assignment.
• The tool's speed is orders of magnitude faster than competing methods.
• AKE is publicly available as a web application for broader accessibility.