Fig 6

(a) Log 2-fold change +/- SE of the expression of the top 30 DE genes across infection type (rhinovirus, coronavirus, coinfection rhinovirus/coronavirus, influenza, and RSV) with respect to negative samples. The top 30 genes were selected as the genes with lowest padj in the LRT comparing the full model (design~ Batch+ParticipantID+Infection_type) to the reduced model, in which the term Infection_type was removed. The significance was not corrected for FDR. (b) Comparison across different viruses of the top immune-related biological processes enriched during infection. Enrichment is quantified by the median zscore of samples that tested positive for each virus (colored dots) and compared to the median zscore across negative samples (black cross). Data underlying panel 6b are zscores reported in S19 Data. (c) Cibersort analysis of the relative abundance of cell types across infection type. Significant pairwise comparisons in cell abundance among viral infections are highlighted in S20 Data. (d) Score factor F₁₀ across the 717 samples calculated with the top 10 genes differentially expressed in the longitudinal analysis and applied to the 6 viral infection groups: negative (Neg), HRV, CoV, CoV+HRV, influenza (FLU), and RSV. Boxplots show the median (red line) and interquartiles (blue lines) of the distribution. Whiskers of boxplots extend to all non-outliers in the distribution. Outliers are shown as red “+”. Panels (a), (c), and (d) are based on data in S17 Data. CoV, coronavirus; LRT, likelihood ratio test; RSV, respiratory syncytial virus.