q2-feature-classifier

q2-feature-classifier provides taxonomic classification of marker-gene amplicon sequences for microbiome composition and functional analyses.

Key Features:

• Machine-learning classifier: Implements a scikit-learn naive Bayes classifier for taxonomic assignment of marker-gene amplicon sequences.
• Alignment-based consensus methods: Supports consensus taxonomy classification using VSEARCH, BLAST+, and SortMeRNA.
• Traditional classifiers: Includes optimized implementations of RDP, BLAST, and UCLUST classifiers.
• Parameter optimization: Emphasizes parameter tuning and provides recommendations to optimize classifier performance for standard conditions.
• Evaluation framework: Integrates the tax-credit evaluation framework for systematic performance assessment of taxonomic classifiers.

Scientific Applications:

• Microbiome taxonomic profiling: Enables accurate taxonomic classification of marker-gene amplicon sequences to characterize microbial community composition and function.

Methodology:

Performs comprehensive evaluation and optimization of machine-learning and alignment-based taxonomic classifiers (scikit-learn naive Bayes, VSEARCH, BLAST+, SortMeRNA, RDP, BLAST, UCLUST) through empirical testing and parameter tuning.