Metax: Cross-Domain Metagenomic Profiling Technology

Reference Number TO 02-99999

Keywords

Metagenomics, Cross-Domain Taxonomic Profiling, Genome Coverage, Microbial Taxa, Contamination Detection, Low Biomass Samples

Invention Novelty

Metax introduces a modern method for metagenomic taxonomic profiling by integrating genome coverage metrics (breadth and depth of coverage) with probabilistic abundance estimation approaches. A key innovation, the Observed-to-Expected Breadth Ratio (OEBR), enhances the accuracy of microbial taxa identification by detecting and minimizing false positives arising from artifact-related irregularities. This feature is especially valuable in low-biomass samples and complex microbial communities, including challenging domains like viruses and archaea.

Value Proposition

Metax improves the reliability of metagenomic taxonomic profiling for clinical diagnostics and ecological studies, especially in low-biomass and host-dominated samples where false positives and false negatives are common, as microbial read counts are inherently low and background noise from various contamination, ambiguous mapping can overwhelm true signals. A key differentiator is coverage-based artifact detection: Metax identifies atypical, localized coverage patterns that often reflect contamination, conserved regions, or reference issues, suppressing spurious taxa while retaining true signals. This leads to more trustworthy pathogen prioritization and more robust, reproducible biomarker discovery across diverse applications. In benchmarks, Metax increases species-level accuracy (higher F1-score) and improves abundance estimates (lower Bray–Curtis distance) compared with widely used profilers.

Metax taxonomy profiling with coverage information: The workflow integrates coverage metrics with probabilistic modeling to estimate species abundance. Artifact signals from contamination, kitome DNA, and shared genetic elements are flagged, while true taxa are confirmed by consistent coverage patterns.

Figure adapted from Deng, Z.-L., Safaei, N., & McHardy, A.C. (2025), bioRxiv, CC BY 4.0.

Technology Description

Metax leverages complete reference genomes to identify taxa rather than marker genes or subsampled sequence signatures. This design improves recall for taxa that are challenging for marker-based or k-mer sketching–based approaches, including small-genome taxa and organisms without well-defined marker sets. Specifically, Metax evaluates genome coverage patterns (breadth, depth, and coverage uniformity) as evidence for true presence and uses an expectation–maximization (EM) algorithm to refine abundance estimates. Coverage-consistency filters then suppress artifact-driven taxa supported by localized, atypical mapping patterns, yielding accurate, contamination-aware cross-domain profiles across sequencing depths and sample types.