CH-Bin

CH-Bin applies a convex-hull distance metric to perform metagenomic contig binning using oligonucleotide composition feature vectors of multiple sizes to improve taxonomic grouping accuracy.

Key Features:

• Convex-hull distance metric: Uses a convex-hull distance to measure geometric relationships among high-dimensional feature vectors for binning.
• Oligonucleotide composition features: Represents contigs as feature vectors derived from oligonucleotide composition of multiple sizes.
• High-dimensional handling: Explicitly addresses issues of the "curse of dimensionality" in composition-based metagenomic binning.
• Comparison to traditional metrics: Provides an alternative to Euclidean and Manhattan distances commonly used for composition- and coverage-based similarity measures.
• Taxonomic binning: Groups contigs into bins corresponding to distinct taxonomic groups.
• Empirical evaluation: Demonstrated improved binning results on both simulated and real datasets.

Scientific Applications:

• Metagenomic contig binning and taxonomic profiling: Enables assignment of assembled contigs to taxonomic groups for downstream analysis.
• Microbial diversity and community structure analysis: Supports characterization of microbial diversity and community composition from environmental metagenomes.
• Environmental microbiology studies: Facilitates investigation of microbial community structure and functional potential in diverse environments.

Methodology:

Contigs are encoded as high-dimensional feature vectors from oligonucleotide composition of multiple sizes, and pairwise similarities are measured using a convex-hull distance metric rather than Euclidean or Manhattan distances for binning.