ngs-shotgun-metagenomics

Use the run_shotgun_metagenomics.py script with a sample sheet and a Kraken2 database path to execute Kraken2 classification, optionally adding Bracken abundance estimation with --run-bracken. Alternatively, run nf-core/taxprofiler via run_nfcore_pipeline.py for a reproducible pipeline execution.

Kraken2 with Bracken provides k-mer-based taxonomic classification and abundance estimation, MetaPhlAn offers marker-gene-based profiling, and HUMAnN quantifies functional pathways and gene families. The Skill supports all three and recommends nf-core/taxprofiler when you want a reproducible multi-tool workflow.

Yes, supplying --host-reference to the backend runner adds a KneadData host-depletion step that writes cleaned FASTQs under host_depletion/ and uses them for downstream Kraken2 and HUMAnN steps. The host reference choice is documented because it can change taxonomic and functional conclusions.

The FASTQ intake package still runs read QC and writes a visualization bundle, but it records the missing database or tool as an explicit blocker in taxonomic_classification_status.json. Summary layers are marked not_available rather than implying profiling succeeded.

You can kick off a preflight test run with nextflow using the test and docker profiles, which validates the pipeline setup. Real analysis still requires appropriate reference databases, and large database downloads should be confirmed for size and destination first.