Skip to main content

Outputs

This page explains the files and directories you should expect after running bff-tools.

Run Directory

Conversion and loading modes create a run-specific project directory. By default, this directory usually has a generated name such as:

beacon_XXXXXXXXXXXXXXX/

You can override the project directory with:

bin/bff-tools vcf -i input.vcf.gz -p param.yaml --projectdir-override my_project

Common Files

OutputCreated byPurpose
log.jsonvcf, tsv, load, fullCaptures resolved arguments, configuration, and parameters for reproducibility
genomicVariationsVcf.json.gzvcf, tsv, fullBFF genomic variation collection generated from VCF-style input
*.log filespipeline stagesStage-specific execution logs for debugging
run_*.sh scriptspipeline stagesRendered shell scripts used to execute the run

Metadata Validation Output

bff-tools validate writes BFF JSON collections to the requested output directory.

Typical files include:

analyses.json
biosamples.json
cohorts.json
datasets.json
individuals.json
runs.json

If genomicVariations validation is enabled, genomic variation output may also be written.

VCF Conversion Output

bff-tools vcf usually writes under:

beacon_*/vcf/

Typical contents include:

  • normalized and annotated VCF intermediates
  • genomicVariationsVcf.json.gz
  • run_vcf2bff.sh
  • run_vcf2bff.log

TSV Conversion Output

bff-tools tsv usually writes under:

beacon_*/tsv/
beacon_*/vcf/

The tsv step first creates a VCF-like intermediate, then continues through the same BFF genomic conversion path.

Browser Output

If bff2html: true is enabled in the parameter file, static browser output is written under:

beacon_*/browser/

This output is useful for local inspection without MongoDB.

MongoDB Loading Output

bff-tools load and bff-tools full usually write MongoDB import logs under:

beacon_*/mongodb/

The actual data is inserted into MongoDB using the configured mongodburi.

Debugging Tips

  • Start with log.json to confirm the resolved configuration and input files.
  • Check the relevant *.log file when a stage fails.
  • Confirm that the reference genome in param.yaml matches your input.
  • For MongoDB issues, inspect the mongodb/ logs and verify mongodburi, mongoimport, and mongosh.