albert-ying/autonomous-lab/kraken2

Kraken2 - Taxonomic Classification

When to Use

Use Kraken2 for taxonomic classification of metagenomic sequencing reads against a reference database.

Standard Workflow

1. Install: conda install -c bioconda kraken2 or brew install kraken2
2. Run classification: kraken2 --db <db_path> --paired <R1.fastq.gz> <R2.fastq.gz> --output <output.txt> --report <report.txt> --threads <N>
3. Key output files:

• output.txt: per-read classification
• report.txt: summary report with read counts per taxon

Key Parameters

• --db: path to Kraken2 database (directory containing hash.k2d, opts.k2d, taxo.k2d)
• --paired: for paired-end reads
• --report: generate summary report
• --confidence: confidence threshold (0-1, default 0)
• --threads: number of threads

Key Decisions

• Confidence threshold affects sensitivity vs specificity tradeoff
• Database choice determines which taxa can be identified