MTR

MTR clusters metagenomic reads at multiple taxonomic ranks to improve taxonomic assignment accuracy and reduce the number of unclassified reads in metagenomic datasets.

Key Features:

• Multi-rank clustering: Generates read clusters at multiple taxonomic ranks and associates each cluster with corresponding taxa at the specific rank.
• Optimization-based selection: Employs a combinatorial optimization algorithm to select a small number of the most informative clusters from each rank.
• Improved assignment vs LCA: Reduces the number of discarded reads compared to Lowest Common Ancestor (LCA) approaches and increases assignments to lower taxonomic ranks.
• Enhanced population characterization: Produces more detailed characterization of metagenome population distributions and microbial community structures.
• Validation: Performance validated on both simulated and real-life metagenomes.
• Implementation: Implemented in Matlab and C++.

Scientific Applications:

• Taxonomic profiling: Refines taxonomic composition analysis of metagenomic datasets by enabling lower-rank assignments.
• Community structure analysis: Enables detailed characterization of microbial community structures and metagenome population distributions.
• Read utilization: Increases utilization of sequencing reads for downstream analyses by reducing unclassified reads.
• Research domains: Supports studies in microbiology, ecology, and evolutionary biology that require accurate taxonomic resolution of environmental sequencing reads.

Methodology:

For each taxonomic rank, MTR generates potential clusters of reads and associates them with taxa at that rank, then applies a combinatorial optimization algorithm to select a small number of the most informative clusters per rank; implementation provided in Matlab and C++.