Skip to content

example.md

Latest commit

 

History

History
62 lines (46 loc) · 1.72 KB

example.md

File metadata and controls

62 lines (46 loc) · 1.72 KB
Raw

Example walkthrough

The tracked example verifies that the complete workflow can start with a PSM table and finish with MSstats, maSigPro, and Mfuzz outputs.

Design

  • two independent-sample groups: Reference and Treatment;
  • six times: 12, 24, 48, 72, 96, and 120;
  • five unique biological replicates per group-time cell;
  • 60 runs and 100 synthetic protein identifiers;
  • six time-matched Treatment-versus-Reference contrasts.

The subset was selected to retain temporal patterns, so it is useful for software testing but biased for biology. All project/sample labels, accessions, and peptide sequences are synthetic. See data/example/README.md for the full de-identification description.

Verify file integrity

Linux:

cd data/example
sha256sum -c SHA256SUMS

macOS:

cd data/example
shasum -a 256 -c SHA256SUMS

Run both stages

docker compose run --rm pdtomsstats \
  Rscript scripts/run_pipeline.R \
  config/example-data.yml \
  config/example-data-timecourse.yml

The main outputs include:

  • results/example/main/deidentified-example.html;
  • results/example/supporting/msstats_all_group_comparisons.csv;
  • one CSV for each planned comparison;
  • results/example/supporting/msstats_workbook_export_objects.rds.

The time-course outputs include:

  • results/example/main/deidentified-example-timecourse.html;
  • masigpro_model_results.csv;
  • masigpro_significant_proteins_by_variable.csv;
  • mfuzz_cluster_membership.csv;
  • mfuzz_cluster_centroids.csv;
  • timecourse_group_difference_effects.csv;
  • a serialized RDS containing the complete time-course objects.

The example deliberately sets run_gsea: false because synthetic protein IDs must not be presented as organism identifiers.