Context-dependent cereal defense and the limits of TRI-cluster transcription as a proxy for deoxynivalenol: A cross-study transcriptomic analysis

Deoxynivalenol (DON) can enter cereal-derived beverages through contaminated ingredients, yet RNA measurements do not quantify toxin burden. We reanalyzed eight public barley, wheat, and Fusarium graminearum transcriptomic series while retaining each study’s design. The analysis included 251 sample-level profiles plus public comparison-level fungal data. At 48 h in barley, eight of 344 eligible Gene Ontology pathways passed discovery-stage bidirectional rules; all eight happened to be upward. Three pathways, including glutathione and aromatic amino acid metabolism, had both purified-DON-associated and infection-associated support, whereas five represented broader infection responses. Seven non-phenylpropanoid pathways remained upward across all 77 pathway-by-contrast tests after limiting each barley gene to one wheat ortholog, although 72 retained false-discovery-rate (FDR) support. In the original exploratory gene screen, eight barley–wheat ortholog pairs were positive across studies; six remained after applying BH FDR < 0.05 to the direct-DON component, and five were also homeolog-concordant. A later barley chemotype dataset yielded no FDR-significant eligible pathway, but its cultivar, chemotype, time, mapping coverage, and sample size limit interpretation. Fungal TRI-cluster transcription increased after FgSfp1 deletion, whereas the source study separately reported 95.4% lower DON yield. This directional comparison was not a matched transcript-toxin calibration. Host pathways and fungal TRI RNA describe context-dependent biological states, not chemical DON concentration. Their possible beverage relevance is limited to upstream hypothesis generation and prioritization of chemical testing.