Beware of misinterpretations in -omics analyses!

A very large proportion of false specific DEG when using few replicates and Venn Diagramms.

Omics analyses represent a crucial step in the search for condition-specific events and often rely on Venn diagrams. In stress experiments for example, the presence of differentially expressed genes exclusively under combined stress conditions is frequently interpreted as evidence of non-additive effects: the effect of combined stresses would differ from the sum of the effects of individual stresses.

In an article published in Nature Plants (Ferraro et al. 2026), to test the relevance of this line of reasoning, the Gnet team and the OGE team with several collaborators at IPS2, and the POPS, PAPPSO and Phenoscope platforms, generated a transcriptomic and proteomic dataset under a control condition, two single stresses, and the combination of both stresses, with 22 biological replicates.

Analysing this dataset, Ferraro et al. demonstrate that identifying specific responses using Venn diagrams generates many false positives. This is explained by the low statistical power inherent in most omics studies that use too few replicates (typically around 3).

Ferraro et al. 2026 thus highlights the need to interpret any claims about the specificity of such omics datasets with caution, and advocate for improvements in statistical modeling to avoid misleading biological interpretations.

Evolution of the number of true and false specific DEGs as a function of the number of replicates. (Left) Venn diagram of DEGs under single and combined stress conditions; (Right) Number of genes correctly identified as specific (blue), falsely identified as specific (red), and falsely identified as differentially ex-pressed (gray) as a function of the number of replicates.

22/06/2026