Skip to content

Configuration templates for a new study

Copy these files into a new project-specific configuration directory; do not edit the templates in place and assume they are analysis-ready.

  • samples.template.tsv records biological identity, experimental factors, sequencing layout, paired read paths, and the final BAM path that will be passed to featureCounts.
  • input_manifest.template.tsv records input roles, provenance, and checksums.
  • analysis_parameters.template.yaml makes study-specific choices explicit.
  • analysis_parameters.agent.template.json provides the same role in a format that can be checked deterministically by the AI-agent validator.

Replace every example row, placeholder path, REVIEW_ME, and REVIEW_FROM_* value. Add columns needed for the actual design, such as sex, time point, tissue, stimulation, or sequencing batch. Remove a column only when it is genuinely irrelevant, not because its value is unknown.

For an AI-assisted local run, place filled copies under config/local/. That directory is ignored by Git so private absolute paths are not published by accident. The templates and configuration-only examples remain tracked.

These general templates do not have the locked schema required by the internship case-study scripts under workflow/. If adapting those scripts, revise their manifest parsing, expected dimensions, factor levels, and validation assertions together with the new study design.