Per-sample data for: Transferability Limits of Published qNet Risk Thresholds Across Sex and Heart Failure in Ventricular Substrates

Simulation output underlying every figure and table of the manuscript "Transferability Limits of Published qNet Risk Thresholds Across Sex and Heart Failure in Ventricular Substrates". Single-cell O'Hara-Rudy simulations, CiPA-optimised variant of Dutta et al. 2017, paced at a basic cycle length of 2,000 ms (0.5 Hz) for 1,000 beats, with biomarkers taken from the final beat. Twelve ventricular substrates are formed from four sex-by-disease phenotypes crossed with three transmural layers. Drug potency was resampled 2,000 times per condition by parametric bootstrap over the IC50 values of Kramer et al. 2013. Contents, 139 files: per_sample/ 96 files, one per layer / drug / concentration / phenotype, 2,000 rows each, one row per uncertainty-quantification sample summaries/ 24 files, per-condition minimum, maximum and median with the change from the matching drug-free baseline traces/ 12 files, full current traces for the drug-free substrates baselines/ 3 files, the twelve drug-free substrates, one row per phenotype thresholds.csv the published per-cell-type qNet decision thresholds reproduce.py recomputes the manuscript's risk classification, drug-free control, charge decomposition and baseline-versus-drug comparison, using only the Python standard library README.md full column definitions, provenance and three cautions LICENSE.txt CC BY 4.0 Compounds: mitoxantrone at 800, 1600, 2400 and 3200 nM, and ribavirin at 2,500, 5,000, 7,500 and 10,000 nM. That is 96 drug conditions and 192,000 drug-exposed beats, plus twelve drug-free baselines. Every label is carried in a filename or a column, never by row or directory position. Concentrations in path names are nanomolar. Files are restricted pending publication of the associated manuscript. Version `1.0.0`