Ait Oumelloul, M.; Saadat, A.; Zanotelli, V.; Tang, S.; Ryan, B.; Chopard, D.; Lawless, D.; Agostini, A.; Nemes-Bokun, I.; Wright, V. J.; Herberg, J.; van der Gaast-de Jongh, C. E.; de Jonge, M.; Sancho-Shimizu, V.; Levin, M.; Agyeman, P. K. A.; Berger, C.; Zamboni, N.; Goetze, S.; Howald, C.; Mannik, K.; Mozun, R.; Schlapbach, L. J.; Froese, D. S.; Fellay, J.; Swiss Pediatric Sepsis Study, ; EUCLIDS Consortium, ; SwissPedHealth Consortium,

Sepsis is a major cause of morbidity and mortality in children, yet biological heterogeneity in host responses has limited progress toward targeted therapies and patient stratification. Multi-omics integration can combine complementary molecular layers to identify coordinated disease programs not captured by individual assays. Here, we integrated genomic, bulk transcriptomic, proteomic and metabolomic data from blood samples of 22 children with culture-confirmed bacterial sepsis enrolled in the Swiss Pediatric Sepsis Study. Multi-Omics Factor Analysis identified a dominant host-response axis reflecting systemic inflammation. This axis was driven primarily by transcriptomic variation and supported by coordinated proteomic and metabolomic signals, including circulating inflammatory mediators and altered amino-acid metabolism. It was associated with C-reactive protein and a severity score proxy. Projection into an independent pediatric sepsis cohort (n = 22) reproduced the inflammatory and severity-related interpretation of this axis. Single-omic projections showed that the integrated signal could be approximated from individual layers, particularly transcriptomics. In three external pediatric whole-blood transcriptomic datasets, the RNA-derived projection separated septic shock from healthy controls and increased across clinical inflammatory syndromes. These findings define a reproducible inflammatory host-response axis in pediatric sepsis and support multi-omics-guided selection of molecular readouts suitable for clinical translation.