Skip to content

WORKFLOW_BENCHMARK.md

Latest commit

 

History

History
29 lines (20 loc) · 2.32 KB

WORKFLOW_BENCHMARK.md

File metadata and controls

29 lines (20 loc) · 2.32 KB

Workflow Benchmark

This repository starts from a published GEO count matrix, so the benchmark below focuses on differential expression and reproducibility workflow quality, not on FASTQ-level alignment or quantification.

Reference workflows

  • DESeq2 vignette: recommends effect-size shrinkage for ranking and visualization.
  • nf-core/rnaseq: exemplifies a mature end-to-end RNA-seq workflow, especially standardized QC and reproducible execution from raw reads.
  • targets user manual: exemplifies dependency-aware skipping and pipeline orchestration.
  • workflowr: exemplifies research provenance via git-aware reporting and session/environment capture.

Current alignment

  • DE effect estimation: raw DESeq2 inference plus apeglm-shrunken log2 fold changes for ranking, MA plotting, and volcano visualization.
  • Robustness: balanced-subset analysis is benchmarked against the full QC-passed cohort, with overlap and concordance written to results/tables/analysis_summary.csv.
  • Reproducibility: renv-pinned environment, deterministic seeds, GitHub Actions rebuilds, and explicit git/session provenance in results/session_info.txt.
  • Artifact validation: tracked tables are rebuilt in CI and compared against committed results; key figures are checked for successful regeneration.

Remaining gaps relative to broader workflows

  • Upstream RNA-seq processing: unlike nf-core/rnaseq, this repo does not perform raw-read QC, alignment, quantification, or MultiQC because the starting point is the GEO count matrix.
  • Pipeline engine: the workflow remains a sequential R-script orchestrator rather than a declarative DAG like targets.
  • Model complexity: the primary DE model remains ~ condition; covariates such as age, sex, viral load, or inferred cell composition are not yet modeled explicitly.

Scope and rationale

  • The codebase is intentionally small and reviewable for a focused secondary analysis.
  • The robustness layer addresses the main biological risk introduced by the balanced-subset design without requiring a full re-architecture.
  • The remaining gaps are explicit, documented, and suitable targets for future extension.