Skip to content

Tutorial

CPTAC lung adenocarcinoma proteomics

The worked tutorial runs against public CPTAC LUAD data, roughly 100 tumour and normal samples, and shows how missingness clusters by tumour/normal status.

Open the notebook on GitHub :material-open-in-new:

It covers:

  1. Loading the CPTAC matrix into the features-by-samples layout mismap-qc expects
  2. The nullity matrix with tumour/normal annotation strips
  3. Per-group completeness
  4. The detection threshold curve
  5. Abundance density of detected against missing values
  6. The validation report from qc()

The data is not in the repository. The notebook's first cell documents where to get it.

Toy data

For a quick look with no download, the repository ships a generator:

uv run make_toy_data.py

That writes data/toy_rnaseq.csv: 80 genes by 30 samples with structured missingness across six groups. demo.py renders the full plot set from it:

uv run demo.py

Both scripts use PEP 723 inline dependencies, so neither needs a virtual environment when run with uv.