Skip to content

Pipeline Decisions

Document the primary analysis pipeline(s), justification for tool selection, and overall data flow. This helps others (and future you) understand the reasoning behind methodological choices.

Pipeline or Tool Used

  • What software, workflow system, or pipeline is being used?
    • Example: Bulk RNA-seq pipeline built with Nextflow; variant annotation using VEP and Varsome API.
  • Is this an existing pipeline, modified version, or custom-built?
    • Example: Custom pipeline integrating Salmon for quantification and DESeq2 for differential expression.

Why Chosen

  • What are the strengths of the selected tool or framework?
    • Example: Nextflow allows modular execution across HPC and local machines.
  • How does the tool align with project needs (e.g., reproducibility, scalability, ease of debugging)?
    • Example: VEP offers transcript-aware variant annotation and integration with known pathogenicity scores.

Key Steps and Data Flow

  • What are the major stages of the pipeline?
    • Example: FASTQ → trimming → alignment → quantification → QC → normalization → statistical testing
  • Where are intermediate and final results stored?
    • Example: Intermediate BAM files stored under /results/alignment/; final DEGs in /results/DESeq2/
  • How are logs, parameters, and environment versions handled?
    • Example: Conda environment YAML versioned with pipeline; logs stored in .log/ and parameters in config.yaml