Derived results and code supporting organ-referenced and annotation-dependent NK-associated RNA contrasts in colorectal liver metastases
Description
Author-generated count aggregates, fixed-marker profiles, patient-level contrasts, descriptive confidence intervals, sensitivity analyses and figure-source tables supporting "Organ-referenced and annotation-dependent NK-associated RNA contrasts in colorectal liver metastases". The package includes analysis and plotting code, official GEO/Zenodo accessions, input specifications and an integrity manifest. Original source matrices, cell/spot expression extracts, VDJ/CDR3 sequence rows and tissue images remain at the source repositories. The included numerical checks compare 75,900 fields across aggregate, regional and patient-summary result tables. These are computational checks of derived results, not independent biological validation or an independent sample count. The README distinguishes runnable checks from archived scripts that require separately acquired source inputs. Detailed descriptions identify the purpose and structure of every file. The analysis examines dependence on organ references, source annotations and measurement choices. It does not establish a common NK-cell state, NK lineage truth, cytotoxic function or causal mechanisms.
Files
Steps to reproduce
Use Python 3.12, install requirements.txt, and run python run_checks.py from the extracted package root. To redraw the two supported v13 summary figures, run python render_v13_figures.py. Read README.md before using archived scripts that depend on excluded source inputs.