Size-aware cross-dataset concordance of public transcriptomic signatures

Pipeline and derived matrices for a three-domain calibration of cross-study agreement among public transcriptomic signatures. The analysis covers 19 human datasets across peripheral blood in COVID-19 (10 datasets), peripheral blood in sepsis (5), and aortic wall in dissection or aneurysm (4), giving 61 pairwise signature comparisons. Key results: median pairwise Jaccard overlap was 0.0185, 0.0759, 0.0624 across the three domains; observed overlap was 2.27, 3.32, 4.15 times the expectation for random gene sets of the same size, with 42 of 61 pairs significant after FDR correction and 12 pairs, all in the COVID-19 domain, at or below chance. Consensus across ten COVID-19 blood datasets retained two genes from a union of 4,225 at 50 per cent agreement and 717 at 20 per cent. Within the COVID-19 domain the excess sits in the whole-blood pairs; the three PBMC pairs fall below chance. The release includes a size-aware concordance measure, a permutation null, consensus analysis across datasets, ground-truth simulations in shared-signal and private-signal regimes, a sweep of the differential-expression rule, and an offline reproduction bundle that regenerates every figure and the chance-corrected tables without network access.