Fig. 6: Network and enrichment analysis of innate cell DEGs, contrasting cures vs. TF.

Panels A, C, E, and G. Significantly DEGs (P_adj < 0.05 and |log₂FC | ≥ 1) were used as input for network analyses. STRING V12.0 was used to construct networks based on co-expression, databases, and experiment terms, with line thickness representing the confidence of the interaction. Enriched categories shown overlaying the networks are those selected from KEGG, GO, or Reactome (one-sided Fisher test with g:SCS-corrected P-values) terms with FDR < 0.05 and strength > 1. The blue border in nodes represents up-regulated genes in TF compared to cures; red borders depict downregulated genes. The intensity of the border color reflects the magnitude of the DE in our dataset. Genes belonging to the Type I IFN cluster are grouped under a blue shade, MHCII in red, and others in gray. Panels B, D, F, H. The same significant DEGs were used for over-representation analyses by gProfiler2. Terms with adjusted P-values ≤ 0.05 (g:SCS method in gProfiler) from the GO biological processes were clustered by term similarity, and the resulting trees were plotted. Groups of nodes are colored by their parent GO term, the size of terminal nodes denotes the number of genes observed in the category, and the color of each node represents the adjusted significance of the observed over-representation.