albert-ying/autonomous-lab/bracken

Bracken - Bayesian Reestimation of Abundance with KrakEN

When to Use

Use Bracken to estimate species/genus/phylum-level abundances from Kraken2 classification results.

Standard Workflow

1. Install: conda install -c bioconda bracken
2. Run after Kraken2: bracken -d <db_path> -i <kraken_report.txt> -o <bracken_output.txt> -r <read_length> -l <level> -t <threshold>
3. Output: abundance table with fraction_total_reads per taxon

Key Parameters

• -d: Kraken2 database path (must contain kmer_distrib files)
• -i: Kraken2 report file
• -o: Output file
• -r: Read length (50, 75, 100, 150, 200, 250, 300)
• -l: Taxonomic level (S=species, G=genus, F=family, O=order, C=class, P=phylum, K=kingdom, D=domain)
• -t: Minimum read threshold (default 0)

Key Decisions

• Read length should match the kmer_distrib files available in the database
• Taxonomic level determines granularity of abundance estimates
• The kmer_distrib files (e.g., database150mers.kmer_distrib) are required