Proteomic Insights from Thrombolytic Ex-Vivo Machine Perfusion in an Extended uDCD Porcine Model

Processed per-sample proteomics data and the analysis code that reproduces every statistic reported in the associated article and its Supplement, together with the code that draws the figures built from those statistics. The targeted 93-protein panel is analyzed with exact-permutation Friedman repeated-measures tests (full enumeration of the within-subject rank permutations; deterministic, no random seed), Durbin-Conover pairwise post-hoc tests, the exact Mann-Whitney U test, the rank-biserial effect size r = 2U/(n1 x n2) - 1, and Benjamini-Hochberg FDR. The whole proteome (8,011 quantified proteins) is analyzed with paired t-tests within the study group and Welch and exact Mann-Whitney tests between groups. Also included are the perfusate blood-gas analysis behind Supplementary Table S19, an independent Python recomputation that asserts agreement with the R output row by row, and a script that re-derives empirically which TMT channels form the normalization reference. Both TMT plexes are referenced to that plex's own six bridging channels. Every script was executed against this archive before deposit and every deposited table was regenerated and matched; the README records what was run and what matched. The archive includes a verification script, verify_deposit.py, that re-runs the entire analysis and compares every regenerated file against the deposited copy, reporting a single pass or fail. The raw mass-spectrometry data are deposited separately with ProteomeXchange via the MassIVE partner repository. Code is licensed MIT; data are licensed CC BY 4.0. Neither license grants patent rights.